Copyright TechTarget - All rights reserved https://cyber.law.harvard.edu/rss/rss.html Techtarget Feed Generator en Wed, 15 Apr 2026 09:06:20 GMT https://www.techtarget.com/searchdatacenter editor@techtarget.com <p>A Linux distribution -- often shortened to <i>Linux distro</i> -- is a complete operating system based on <a href="https://www.techtarget.com/searchdatacenter/definition/Linux-operating-system">Linux</a>. It is packaged with other components and software to make it fully usable out of the box. Because Linux is very modular, different distros will focus on different target use cases.</p> <p>Linux distributions, which are based on the Linux <a href="https://www.techtarget.com/searchdatacenter/definition/kernel">kernel</a>, are often easier for users to deploy than the <a href="https://www.techtarget.com/whatis/definition/open-source">open source</a> version of Linux. The Linux kernel is released as <a href="https://www.techtarget.com/searchapparchitecture/definition/source-code">source code</a>, which needs to be compiled before it is used. The kernel also does not contain many of the elements that a user might need, such as installation programs, a graphical user interface, management tools and additional software, such as a <a href="https://www.techtarget.com/searchitoperations/tip/Hyper-V-vs-KVM-Select-the-right-hypervisor-for-your-IT-needs">KVM hypervisor</a>. A distro takes all these elements and brings them together into a single release.</p> <section class="section main-article-chapter" data-menu-title="Linux distribution types"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Linux distribution types</h2> <p>Hundreds of Linux distributions are available today, and each targets specific users or systems, such as desktops, servers, mobile devices or embedded internet-of-things (<a href="https://www.techtarget.com/iotagenda/definition/Internet-of-Things-IoT">IoT</a>) devices. A major appeal of Linux is its adaptability. The same Linux kernel can run everything from a small single central processing unit (CPU) temperature sensor, to a modern desktop environment, to a ten-thousand-CPU <a href="https://www.techtarget.com/whatis/definition/supercomputer">supercomputer</a>. The differing needs of these systems can be met by using different Linux distros.</p> <p>Some distributions -- such as <a href="https://www.techtarget.com/searchdatacenter/definition/Fedora">Fedora</a> and Red Hat Enterprise Linux from Red Hat, openSUSE from <a href="https://www.techtarget.com/searchdatacenter/definition/SuSE">SUSE</a>, <a href="https://www.techtarget.com/searchdatacenter/definition/Ubuntu">Ubuntu</a> from Canonical, and Oracle Linux from Oracle -- are commercial, while others -- such as <a href="https://www.techtarget.com/searchdatacenter/definition/Debian">Debian</a> and Slackware -- are community-developed. Some commercial distributions, for example, those from Red Hat and Oracle, charge users for services, such as support or custom development, although open-source licensing prohibits charging for the open-source software itself.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/popular_linux_distributions-f.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/popular_linux_distributions-f_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/popular_linux_distributions-f_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/popular_linux_distributions-f.png 1280w" alt="Chart of popular Linux distributions." height="353" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>Here are some popular Linux distributions among the hundreds that exist. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> </section> <section class="section main-article-chapter" data-menu-title="Community vs. commercial Linux distributions"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Community vs. commercial Linux distributions</h2> <p>One of the most important decisions when choosing a Linux distribution is whether to use a community (free) or commercial (paid) version. Because Linux has open-source software at its core, every commercial company that has a paid distro must also have a free version, often called <em>community version</em>. It is tempting to think that the free version can be used as a money-saving tactic, but for many organizations, the <a href="https://www.techtarget.com/searchcio/tip/5-open-source-software-problems-and-how-to-manage-them">trade-offs</a> are not worth it.</p> <p>The primary differentiator between community and commercial Linux distributions is the level of service provided by the vendor. The free community distro is often released <i>as is</i> with no additional provided support or guarantees. The only way to get help is by posting to a forum and asking other users for assistance. The paid versions, however, often come with service contracts and response time guarantees. The <a href="https://www.techtarget.com/searchdatacenter/tip/Top-5-options-for-Linux-certifications">support staff</a> can work with the developers directly and may be able to get custom fixes implemented.</p> <p>The value of service contracts becomes clear when you consider how important Linux is to many organizations. For example, if the 100-node <a href="https://www.techtarget.com/whatis/definition/Web-server">web server</a> for a shopping platform were to have a problem that causes it to lose 1% of sales -- this would be a critical issue that could be difficult to diagnose. Direct communication with the distro maintainer could help resolve it much faster.</p> </section> <section class="section main-article-chapter" data-menu-title="Long-term support distros"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Long-term support distros</h2> <p>Linux is constantly being updated. While this can be good from a user point of view, for enterprise deployments, <a href="https://www.techtarget.com/searchenterprisedesktop/tip/Top-5-enterprise-Linux-distributions-to-consider-adopting">stability is more important</a>. Some Linux distros are called <em>long-term support (LTS)</em> to denote that they will stay stable for many years.</p> <p>LTS distros will generally not add new features or improvements. They will update for bug fixes and security <a href="https://www.techtarget.com/whatis/feature/5-reasons-software-updates-are-important">updates</a>. This is beneficial as it makes them more reliable for deployment as servers. Users can be reasonably sure that systems which run on LTS will operate the same for as long as it is supported.</p> <p>LTS distros also have very clearly defined lifecycles of many years. This allows for long-term planning and allows the same system to be used for a prolonged time without needing to reengineer it when a new version of the operating system comes out.</p> </section> <section class="section main-article-chapter" data-menu-title="Linux distro packages"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Linux distro packages</h2> <p>Generally, Linux distributions consist of what are called software packages. These packages contain specific files, applications or services. For example, a package could be a collection of fonts, web browsers or <a href="https://www.techtarget.com/searchsoftwarequality/definition/development-environment">development environments</a>. A single Linux distribution could contain thousands of software packages. Many of these packages are not preinstalled.</p> <p>A Linux distribution also includes a package management system, or packet manager, which is used to install, uninstall and manage software packages. These systems also allow for package searches, automatic software upgrades and verification that all package dependencies are fulfilled. Examples of package managers include Red Hat Package Manager, Yellowdog Updater Modified and Advanced Packaging Tool.</p> <p>These package repositories are often not the same between distros. The software available out-of-the-box for Ubuntu might not be the same as for Red Hat, for example.</p> </section> <section class="section main-article-chapter" data-menu-title="Open source development"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Open source development</h2> <p>Linux is founded upon open source software development. Linux distributions emerged under the copyleft stipulations of the Free Software Foundation, which originated the GNU General Public License (<a href="https://www.techtarget.com/searchdatacenter/definition/GNU-General-Public-License-GNU-GPL-or-simply-GPL">GPL</a>). Copyleft dictates that any software taken for free and altered must be consequently distributed for free.</p> <p>So if a developer uses Linux, or <a href="https://www.techtarget.com/searchdatacenter/definition/GNU-Linux">GNU</a> components, to create a new version of Linux, the new version must be free. Commercial Linux vendors such as Red Hat typically generate revenue from services and tools rather than the Linux package itself.</p> </section> <section class="section main-article-chapter" data-menu-title="Linux distro families"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Linux distro families</h2> <p>Because most Linux distributions are released as open source, it is possible to take one distro as the core and release it as a new one with modifications. This means that many distros are derivatives of other ones, which are called forks. This creates families or groups of distros that share common <a href="https://www.techtarget.com/whatis/definition/component">components</a> but might be targeting different audiences. The term upstream or downstream can be used to denote if a particular distro receives or contributes <a href="https://www.techtarget.com/whatis/definition/code">code</a> to another distro.</p> <ul class="default-list"> <li><b>Debian Linux.</b> This is one of the oldest and most open Linux distributions. It serves as the base for many of the most popular distros in use today. Ubuntu Linux by Canonical is one of the most recognizable with versions for community, licensed, desktops, servers and long-term support. Other popular Debian-based desktop distros include Linux Mint, Pop-OS, <a href="https://www.techtarget.com/searchitoperations/tip/What-is-a-Raspberry-Pi-used-for">Raspberry Pi</a> OS and Kali Linux.</li> <li><b>Red Hat. </b>Red Hat is a company focused on making enterprise grade Linux. Fedora is its community distribution and Red Hat Enterprise Linux (<a href="https://www.techtarget.com/searchdatacenter/definition/Red-Hat-Enterprise-Linux-RHEL">RHEL</a>), its paid support distro. Some downstream RHEL distros include Oracle Linux, Rocky Linux and AlmaLinux.</li> <li><b>Arch Linux. </b>This is a popular upstream distro that focuses on minimalism with additional packages being available on demand.</li> <li><b>SUSE. </b>This is another major enterprise Linux vendor. It has commercial and community versions of SUSE Linux.</li> </ul> <p>Google's <a href="https://www.techtarget.com/searchmobilecomputing/definition/Android-OS">Android</a><b> </b>operating system is based on Linux and can be considered a very specialized distro.</p> <p>One common misconception is that Apple's <a href="https://www.techtarget.com/whatis/definition/Mac-OS">macOS</a> is based on Linux. It is not a Linux distro, instead, macOS is based on <a href="https://www.techtarget.com/searchdatacenter/definition/Unix">UNIX</a>.</p> <div class="youtube-iframe-container"> <iframe id="ytplayer-0" src="https://www.youtube.com/embed/fdizzTv4aNA?si=Cvts3oCDIUU1AQDi?autoplay=0&amp;modestbranding=1&amp;rel=0&amp;widget_referrer=null&amp;enablejsapi=1&amp;origin=https://www.techtarget.com" type="text/html" height="360" width="640" frameborder="0"></iframe> </div> <p></p> </section> A Linux distribution -- often shortened to 'Linux distro' -- is a complete operating system based on Linux. https://cdn.ttgtmedia.com/visuals/digdeeper/2.jpg https://www.techtarget.com/searchdatacenter/definition/Linux-distros-Linux-distribution Fri, 31 Oct 2025 15:01:00 GMT <p>A configuration management database (CMDB) is a file -- usually in the form of a standardized <a href="https://www.techtarget.com/searchdatamanagement/definition/database">database</a> -- that contains all relevant information about the hardware and software components used in an organization's IT services and the relationships among those components. A CMDB stores information that provides an organized view of configuration data and a means of examining that data from any desired perspective.</p> <p>As IT infrastructure becomes <a href="https://www.computerweekly.com/blog/Data-Matters/The-importance-of-simplifying-IT-enterprise-capability-without-the-complexity">more complex</a>, the importance of tracking and understanding the information in the IT environment increases. The use of CMDBs is a best practice for IT teams and leaders who need to identify and verify each component of their infrastructure to better manage and improve it.</p> <section class="section main-article-chapter" data-menu-title="How CMDBs work and why they are important"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How CMDBs work and why they are important</h2> <p>In the context of a CMDB, components of an information system are referred to as <i>configuration items</i> (CIs). CIs can be any conceivable IT components, including software, hardware, documentation and personnel. They can also indicate the way in which each CI is configured and any relationship or dependencies among them. <a href="https://www.techtarget.com/searchitoperations/definition/configuration-management-CM">Configuration management</a> processes seek to specify, control and track CIs and any changes made to them in a comprehensive, systematic fashion.</p> <div class="youtube-iframe-container"> <iframe id="ytplayer-0" src="https://www.youtube.com/embed/66gpOMI2m4Y?autoplay=0&amp;modestbranding=1&amp;rel=0&amp;widget_referrer=null&amp;enablejsapi=1&amp;origin=https://www.techtarget.com" type="text/html" height="360" width="640" frameborder="0"></iframe> </div> <p>CMDBs capture CI attributes, including importance, ownership and identification code. A CMDB also provides details about CI relationships and dependencies; this makes it a powerful tool if used correctly. As a business enters more CIs into the system, the CMDB becomes a stronger resource to predict changes in the organization. For example, <a href="https://www.computerweekly.com/opinion/One-year-on-from-the-CrowdStrike-outageWhat-have-we-learned">if an outage occurs</a>, IT can understand from CI data which systems are affected.</p> <p>A CMDB can be used for many activities besides capturing CI data, including the following:</p> <ul class="default-list"> <li>Performing problem management.</li> <li>Conducting <a href="https://www.techtarget.com/searchitoperations/definition/root-cause-analysis">root cause analysis</a>.</li> <li>Identifying potential vulnerabilities.</li> <li>Complying with regulatory metrics.</li> <li>Investigating workflows.</li> <li>Reducing <a href="https://www.techtarget.com/searchdatabackup/feature/The-cost-of-downtime-and-how-businesses-can-avoid-it">downtime</a>.</li> <li>Enhancing service delivery.</li> <li>Optimizing business services.</li> <li>Tracking software licenses.</li> <li>Capturing real-time data on potential performance issues.</li> </ul> <p>The CMDB connects to virtually every element in the IT infrastructure. It <a href="https://www.techtarget.com/searchitoperations/feature/Configuration-management-vs-asset-management-simplified">provides asset management, as well as configuration data</a>, for system and network administration and security management. CMDB data is typically presented on a dashboard display.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/where_a_configuration_management_database_fits_in_it_infrastructure-f.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/where_a_configuration_management_database_fits_in_it_infrastructure-f_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/where_a_configuration_management_database_fits_in_it_infrastructure-f_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/where_a_configuration_management_database_fits_in_it_infrastructure-f.png 1280w" alt="A diagram showing how a configuration management database fits in IT infrastructure." height="403" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>A configuration management database contains information about all the hardware and software components in an organization's IT infrastructure. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> </section> <section class="section main-article-chapter" data-menu-title="Features of a CMDB"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Features of a CMDB</h2> <p>CMDBs are centralized repositories that capture and store data about IT assets, their configurations, and relationships. Among a CMDB's key features are workspace, data acquisition and integration, visualization and reporting. Here's a description of core CMDB features:</p> <ul class="default-list"> <li><b>CMDB workspace</b>. Provides a resource for managing and viewing CIs and how they interact.</li> <li><strong>Data acquisition and integration. This</strong><strong> </strong><strong>process</strong><strong> </strong><strong>captures</strong><strong> </strong><strong>and</strong><strong> </strong><strong>integrates</strong><strong> </strong>data from multiple sources, such as sensors, creating a total view of IT assets.</li> <li><b>Mapping of relationships.</b> The CMDB presents visually how different CIs interact and depend on each other; this facilitates operational analysis and change management.</li> <li><b>Visualization and reporting</b>. Prepares and presents detailed maps and diagrams of how CIs interact, helping the business understand how it uses CI relationships.</li> <li><b>Centralized asset management</b><b>.</b> Provides a single unified view of all IT assets, accommodating a single view of the truth.</li> <li><b>Compliance. </b>Data gathered from a CMDB can show how a system complies with specific standards and <a href="https://www.techtarget.com/searchcio/definition/regulatory-compliance">regulations</a>.</li> <li><b>Access controls.</b> These govern access to the CMDB and detail how access is managed throughout the infrastructure.</li> <li><b>Lifecycle management.</b> CMDB data can be used to ensure all assets are being managed in line with their expected lifecycles.</li> <li><b>Root cause analysis.</b> CMDB data may be used as part of a root cause analysis, especially after a service disruption</li> <li><b>Risk and change management.</b> CMDB data can support risk assessments and change management activities.</li> <li><b>Incident and problem management.</b> Armed with CMDB data, technical staff responding to an incident or technical problem can examine the asset database for insights.</li> </ul> <p>IT teams use CMDB features to manage their technology and networking infrastructures, improve resource visibility and facilitate IT activities such as change and incident management.</p> </section> <section class="section main-article-chapter" data-menu-title="Who needs CMDBs?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Who needs CMDBs?</h2> <p>IT organizations need CMDBs to capture information about the CIs. CMDBs can be paired with asset management systems to identify all elements in an IT infrastructure. CMDBs build on asset inventories, providing information on the relationships among CIs.</p> <p>Organizations use the CMDB to predict changes that can affect IT systems, which systems will be affected and how. IT administrators can also use CMDB data to identify when it's appropriate or necessary to replace a device or other asset.</p> </section> <section class="section main-article-chapter" data-menu-title="Advantages of a CMDB"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Advantages of a CMDB</h2> <p>CMDBs provide various benefits, including the following:</p> <ul class="default-list"> <li><b>Centralized view of data.</b> This capability gives IT administrators more control over the IT infrastructure. Admins can get data on each component in an IT infrastructure -- like a storage device or an application running on a server. This helps with planning, managing and maintaining the entire infrastructure. It also lowers the incidence of administrative and management errors, helps to ensure regulatory compliance, and increases security.</li> <li><b>Cost savings.</b> CMDBs help IT managers spot ways to eliminate unnecessary or redundant IT resources and their associated costs.</li> <li><b>Data integration.</b> CMDBs let admins integrate data from various vendors' software, reconcile that data, identify any inconsistencies in the database and ensure all data is synchronized. A CMDB system can also integrate other configuration-related processes, such as <a href="https://www.techtarget.com/searchcio/definition/change-management">change management</a> and <a href="https://www.techtarget.com/searchitoperations/definition/IT-incident-management">incident management</a>.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="Challenges of a CMDB"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Challenges of a CMDB</h2> <p>A CMDB can also present several challenges. A particularly difficult issue is organizational: to convince the business of the benefits of a CMDB and then to use the system properly once implemented.</p> <p>Other challenges include:</p> <ul class="default-list"> <li><b>Importing relevant data.</b> This can be a tedious task. Admins must input a wealth of information about each IT asset, including financial information, upgrade history and performance profile. Modern CMDB tools offer enhanced discovery capabilities, enabling the tool to find and profile CIs automatically. However, this data doesn't always come from the same source. In theory, a process called <a href="https://www.techtarget.com/searchbusinessanalytics/news/252507049/Modern-data-strategy-includes-cloud-domain-federation"><i>data federation</i></a> brings together data from disparate locations to prevent IT from replacing or eliminating other data systems. In practice, data is dispersed across sources that aren't well integrated, which prevents IT managers from federating data.</li> <li><b>Updating and maintaining CMDBs.</b> Over time, IT administrators must <a href="https://www.techtarget.com/searchitoperations/tip/Maintain-CMDB-data-integrity-for-automated-disaster-recovery">regularly review, update and maintain CMDB data</a>. A CMDB can fail if admins don't update the data, in which case it becomes stale and unusable.</li> </ul> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/itops-cmdb_challenges-f.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/itops-cmdb_challenges-f_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/itops-cmdb_challenges-f_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/itops-cmdb_challenges-f.png 1280w" alt="A chart detailing challenges of a configuration management database system in the areas of data entry, data lifecycles, data utilization and data storage and protection." height="342" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>Data integrity is the cornerstone of a good configuration management database system. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <p></p> </section> <section class="section main-article-chapter" data-menu-title="CMDB best practices"> <h2 class="section-title"><i class="icon" data-icon="1"></i>CMDB best practices</h2> <p>Several activities can be considered best practices when planning, implementing and managing a CMDB. As with any technology implementation, careful planning and alignment with business requirements are essential to a successful project. Additional best practices:</p> <ol class="default-list"> <li><b>Define operating objectives.</b> Determine the goals of a CMDB, such as enhancing asset tracking, change management and incident response.</li> <li><b>Secure management approval and funding.</b> This is essential to ensure the CMDB initiative is fully supported and funded.</li> <li><b>Identify primary configuration items (CIs).</b> Launch the CMDB with critical assets, services, and operating relationships.</li> <li><b>Keep data accurate and current.</b> Minimize the likelihood of data inconsistencies, duplicates and outdated data by regularly reviewing, validating and reconciling data.</li> <li><b>Use automation when possible.</b> If the CMDB offers automation tools, use them for activities such as CI discovery and updating to minimize errors.</li> <li><b>Governance and access controls.</b> Manage the CMDB by defining roles and <a href="https://www.techtarget.com/searchsecurity/tip/Types-of-access-control">access permissions</a>.</li> <li><b>Integrate CMDB with ITSM and related assets.</b> For optimal operational efficiency, integrate the CMDB with incident, change, and service management activities.</li> <li><b>Dependency and relationship mapping.</b> Identifying how various assets work with each other is essential; capturing those relationships and dependencies eases incident response and problem resolution.</li> <li><b>Testing and performance analysis.</b> Periodic CMDB tests can ensure that it is performing properly and tee up the resolution of potential issues.</li> <li><b>Reviewing, auditing and monitoring.</b> Make sure the CMDB is reviewed periodically for completeness, accuracy, and <a href="https://www.techtarget.com/searchcio/feature/What-is-IT-business-alignment-and-why-is-it-important">alignment with business needs</a>.</li> <li><b>Training and documentation.</b> Provide training and any descriptive content on the CMDB to users to ensure the CMDB is leveraged effectively.</li> <li><b>Continuous improvement.</b> The CMDB is a living resource. Set performance metrics, gather feedback, and refine CMDB operations for <a href="https://www.techtarget.com/searcherp/definition/kaizen-or-continuous-improvement">continuous improvement</a>.</li> </ol> </section> <section class="section main-article-chapter" data-menu-title="Evolution of the CMDB"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Evolution of the CMDB</h2> <p>As a <a href="https://www.techtarget.com/searchnetworking/post/Network-automation-success-begins-with-a-source-of-truth">single source of truth</a> of configuration information for IT assets, a CMDB facilitates monitoring of assets and dependencies, making upgrades deployment of new services easier. For example, CMDB data can help identify which servers run an older operating system (OS) version and how <a href="https://www.computerweekly.com/news/366627608/Current-approaches-to-patching-unsustainable-report-says">patches might alter security</a> and performance.</p> <p>Organizations can track and enforce CMDB information over time, which can improve security and compliance and reduce risks. CMDBs also play a central role in <a href="https://www.techtarget.com/searchitoperations/tip/Enable-automated-failover-with-a-highly-available-CMDB">automated failover</a> and disaster recovery activities.</p> <p>The term <i>configuration management</i> continues to expand its meaning to reflect the increased use of <a href="https://www.techtarget.com/searchitoperations/feature/The-evolution-and-history-of-software-configuration-management">software-based configurations and interactions</a>: scripting the configuration of a software stack, container management and <a href="https://www.techtarget.com/searchitoperations/definition/Google-Kubernetes">Kubernetes</a>, automation down to the code level, and cloud resources and provisioning.</p> <p>The DevOps universe of technologies and practices, including containers, microservices, <a href="https://www.techtarget.com/searchitoperations/tip/Infrastructure-as-code-principles-How-IaC-works-and-how-to-use-it">infrastructure as code</a>, source control, package management and release automation, has changed what it means to map and track asset configurations and dependencies. Machine learning and AI promise to predict the impact of undesirable results more quickly and accurately from configuration changes and their propagation.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/itops-benefits_of_software_configuration_management-f.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/itops-benefits_of_software_configuration_management-f_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/itops-benefits_of_software_configuration_management-f_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/itops-benefits_of_software_configuration_management-f.png 1280w" alt="An infographic detailing the benefits of software configuration management" height="430" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>Software configuration management provides several benefits to organizations seeking more control over their software development process, from source code to APIs to change requests. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <p>Configuration management for tracking configuration changes in physical and digital assets remains essential. Organizations still must understand the landscape of their IT infrastructure resources and how the interplay of those resources supports business objectives.</p> <p>CMDBs have evolved to more closely align with IT service management (<a href="https://www.techtarget.com/searchitoperations/definition/ITSM">ITSM</a>) and reporting capabilities, as well as the cloud and distributed infrastructure. Many CMDBs integrate with IT asset management (<a href="https://www.techtarget.com/searchcio/definition/IT-asset-management-information-technology-asset-management">ITAM</a>) platforms, which are similar information repositories about IT assets that support change management. CMDBs can also be used to store such information themselves.</p> </section> <section class="section main-article-chapter" data-menu-title="CMDBs and ITIL"> <h2 class="section-title"><i class="icon" data-icon="1"></i>CMDBs and ITIL</h2> <p>The IT Infrastructure Library <a target="_blank" href="https://www.itlibrary.org/" rel="noopener">service management framework</a> includes specifications for configuration management, although adoption of the <a href="https://www.techtarget.com/searchdatacenter/definition/ITIL">ITIL</a> framework isn't a prerequisite for configuration management. According to ITIL specifications, the four major aspects of configuration management are:</p> <ol class="default-list"> <li><b>Discovery.</b> Identify CIs to be included in the CMDB.</li> <li><b>Security.</b> Control data to ensure only authorized individuals can change it.</li> <li><b>Reporting.</b> Maintain status, ensuring that the status of any CI is recorded and updated consistently.</li> <li><b>Auditing.</b> Verify accuracy through <a href="https://www.techtarget.com/searchcio/tip/Prep-a-compliance-audit-checklist-that-auditors-want-to-see">audits</a> and reviews of the data.</li> </ol> <p>Prior ITIL versions introduced and expanded the importance of configuration management, which is designed to capture details of all configuration items as part of ITSM activities. The most recent ITIL release, <a href="https://www.techtarget.com/searchitoperations/opinion/ITIL-4-framework-brings-long-awaited-flexibility-to-ITSM">ITIL v4</a> (2019), defined an IT operations model for delivering products and services. It plays a role in the overall business strategy.</p> </section> <section class="section main-article-chapter" data-menu-title="CMDBs vs. ITAM"> <h2 class="section-title"><i class="icon" data-icon="1"></i>CMDBs vs. ITAM</h2> <p>There is functional overlap between CMDBs and ITAM platforms for change management. Their capabilities are also increasingly integrated into broader service management frameworks. However, they are different tools <a href="https://www.techtarget.com/searchdatacenter/tip/Figure-out-the-differences-of-asset-management-vs-CMDB">used for different purposes</a>.</p> <p>ITAM tools track asset data, such as hardware and software details, across the entire <a href="https://www.techtarget.com/searchdatacenter/tip/IT-asset-retirement-in-the-data-center">asset lifecycle</a>. That data tends to be more static than the dynamic activities a CMDB tracks: acquisition and procurement, operation, change management, maintenance and disposal.</p> <p>ITAM data includes configuration information. It also tracks costs at each lifecycle stage, such as purchasing and licensing, service, support and depreciation. Asset management benefits include better asset utilization and proactive asset compliance and security auditing. Improved asset visibility also leads to faster and more accurate business decision-making.</p> <p>ITAM tools are typically used to achieve business-oriented goals, such as making and reviewing decisions through an infrastructure asset lifecycle. Configuration management tools are better suited for service-oriented goals, helping IT staff understand dependencies so they can plan and maintain IT services. Change management is an important CMDB activity.</p> <p>ITAM and CMDBs are not mutually exclusive. For example, an application server is an IT asset with financial value that depreciates over time. It also requires maintenance and can incorporate operational information, such as <a href="https://www.techtarget.com/searchitoperations/tip/Manage-your-IT-service-contracts-to-save-money">service agreements</a>, that are not part of a CMDB. That server is also a CI, and information about it can be tracked and managed through a CMDB, including its installed OS and software, server setup and firmware versions. The CMDB could reveal how changes to the server's configuration state might affect performance, stability and security; this is called an <i>impact analysis.</i></p> </section> <section class="section main-article-chapter" data-menu-title="CMDB vendors and tools"> <h2 class="section-title"><i class="icon" data-icon="1"></i>CMDB vendors and tools</h2> <p>General CMDB capabilities include the following:</p> <ul class="default-list"> <li>Discover and assess the CI of IT assets.</li> <li>Automatically update CMDB entries when an asset is changed or updated.</li> <li>Map dependencies between assets and CIs.</li> <li>Simulate or predict the effect of a change to CIs.</li> <li>Audit CMDB records for security and compliance initiatives.</li> <li>Ensure compliance with relevant standards and regulations.</li> </ul> <p>Many configuration management, asset management and CMDB tools are available for enterprises of various sizes and needs. Here are some available tools:</p> <ul class="default-list"> <li>AlgoSec.</li> <li>Atomicwork.</li> <li>BMC Helix CMDB.</li> <li>Broadcom CA Service Management.</li> <li>Canfigure.</li> <li><a href="https://www.techtarget.com/searchnetworking/news/252516203/Device42-adds-intelligence-to-IT-discovery-asset-management">Device42</a>.</li> <li>Freshservice.</li> <li>GLPI.</li> <li>IBM Control Desk.</li> <li>IBM Tivoli Change and Configuration Management Database.</li> <li>InvGate Insight.</li> <li>ManageEngine AssetExplorer.</li> <li>Microsoft System Center Service Manager.</li> <li>OpenText Universal Discovery and Universal CMDB.</li> <li><a href="https://www.techtarget.com/searchitoperations/feature/ServiceNow-Configuration-Management-Database">ServiceNow CMDB</a>.</li> <li><a href="https://www.techtarget.com/searchitoperations/news/252522170/ServiceNow-ITSM-users-recharge-workflows-with-familiar-tools">ServiceNow ITSM Enhancer</a>.</li> <li><a href="https://www.techtarget.com/searchitoperations/tip/How-and-why-to-add-SolarWinds-modules">SolarWinds Service Desk</a>.</li> <li><a href="https://www.techtarget.com/searchitoperations/news/252465726/SysAid-launches-Automate-Joe-for-ITSM-platform-automation">SysAid Technologies</a>.</li> <li>TOPdesk.</li> <li>Virima.</li> </ul> <p>Integrated and third-party tools are also available to supplement a CMDB. Examples include the following:</p> <ul class="default-list"> <li><b>ITSM tools.</b> They can integrate with CMDBs and often incorporate CMDB capabilities of their own. Many ITSM vendors offer standalone CMDBs as well. Tools from a single vendor may offer integration advantages but less so for users of third-party CMDBs.</li> <li><b>Automated discovery and change management tools. </b>They automatically generate and update data to capture the state of the IT environment. However, while discovery tools enable IT to take a more hands-off approach to configuration management, they don't eliminate the need for manual entry. For example, some details such as the hardware's purchase date, price and due date of the next renewal of service may require manual entry.</li> <li><b>IT operations analytics tools. </b>They can integrate with CMDBs. These tools can analyze the established configuration of each server, compare possible changes against an existing <a href="https://www.techtarget.com/searchcio/definition/benchmark">benchmark</a> and alert IT managers to unexpected or disallowed changes to a configuration for examination and remediation.</li> <li><b>Data management tools. </b>They can address data federation by taking all IT data from a variety of sources and automatically storing it in a CMDB. Such tools increase the accuracy of an enterprise's CMDB data.</li> <li><b>Unified endpoint management and software asset management tools. </b>These<b> </b>are used as data sources for a CMDB to provide visibility for devices in their control.</li> </ul> <p><i>Find out more about the </i><a href="https://www.techtarget.com/searchitoperations/tip/How-change-management-and-configuration-management-differ-in-IT"><i>relationship between change management and configuration management</i></a></p> </section> A configuration management database (CMDB) is a file -- usually in the form of a standardized database -- that contains all relevant information about the hardware and software components used in an organization's IT services and the relationships among those components. https://cdn.ttgtmedia.com/visuals/digdeeper/4.jpg https://www.techtarget.com/searchdatacenter/definition/configuration-management-database Thu, 23 Oct 2025 09:00:00 GMT <section class="section main-article-chapter" data-menu-title="What is data center infrastructure efficiency?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What is data center infrastructure efficiency?</h2> <p>Data center infrastructure efficiency (DCiE) is a metric used to determine the energy efficiency of a data center by measuring what percentage of total facility power is consumed by IT equipment.</p> <p>DCiE was developed by members of The Green Grid, an industry group focused on&nbsp;<a href="https://www.techtarget.com/searchdatacenter/tip/Four-ways-to-reduce-data-center-power-consumption">data center energy efficiency.</a></p> </section> <section class="section main-article-chapter" data-menu-title="How to calculate DCiE"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How to calculate DCiE</h2> <p>DCiE, which is expressed as a percentage, is calculated by dividing IT equipment power by total facility power.</p> <p>DCiE = IT Equipment Power / Total Facility Power x 100%</p> <p>DCIE is the reciprocal of power usage effectiveness (<a href="https://www.techtarget.com/searchdatacenter/definition/power-usage-effectiveness-PUE">PUE</a>). PUE is defined as the total facility power divided by the IT equipment power.</p> <p>DCiE = 1 / PUE</p> </section> <section class="section main-article-chapter" data-menu-title="How to determine energy use"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How to determine energy use</h2> <p>Admins can monitor and measure the total facility power and IT equipment load.</p> <h3>Total facility power measurement</h3> <p>Measure <a href="https://www.techtarget.com/searchdatacenter/tip/How-much-energy-do-data-centers-consume">energy use</a> at or near the facility's utility meter. If the data center is in a mixed-use facility or office building, measure only at the meter that is powering the data center. If the data center is not on a separate utility meter, estimate the amount of power being consumed by the non-data center portion of the building and remove it from the equation.</p> <h3>IT equipment load measurement</h3> <p>Measure the IT equipment load, which should be measured after power conversion, switching and conditioning are completed. According to The Green Grid guidelines, the most likely measurement point would be at the output of the computer room&nbsp;<a href="https://www.techtarget.com/searchdatacenter/definition/power-distribution-unit-PDU">power distribution units</a>. This measurement should represent the total power delivered to the server racks in the data center.</p> <div class="youtube-iframe-container"> <iframe id="ytplayer-0" src="https://www.youtube.com/embed/ucmnHYCawyA?autoplay=0&amp;modestbranding=1&amp;rel=0&amp;widget_referrer=null&amp;enablejsapi=1&amp;origin=https://www.techtarget.com" type="text/html" height="360" width="640" frameborder="0"></iframe> </div> </section> <section class="section main-article-chapter" data-menu-title="Current standards and measurement levels"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Current standards and measurement levels</h2> <p style="font-size: 16px;">The Green Grid specifies three different levels for measuring energy usage for both PUE and DCiE calculations.</p> <ul style="font-size: 16px;" class="default-list"> <li><b>Level 1 (Basic)</b>: Monthly/weekly measurements.</li> <li><b>Level 2 (Intermediate)</b>: Hourly/daily measurements.</li> <li><b>Level 3 (Advanced)</b>: Continuous measurements taken in intervals of 15 minutes or less.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="Modern context and limitations"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Modern context and limitations</h2> <p style="font-size: 16px;">As of 2024-2025, while DCiE and PUE remain key metrics for measuring data center efficiency, industry experts recognize their limitations. The Green Grid now recommends using additional complementary metrics, such as:</p> <ul style="font-size: 16px;" class="default-list"> <li>Advanced Mechanical Load Component (AMLC)</li> <li>IT Equipment Work Capacity (ITWC)</li> <li>Total Usage Effectiveness (TUE)</li> <li>Water Usage Effectiveness (WUE)</li> </ul> <p style="font-size: 16px;">These additional metrics provide a more comprehensive picture of data center sustainability and efficiency, especially as the industry faces growing demands from AI workloads and increasing focus on environmental responsibility.</p> </section> Data center infrastructure efficiency (DCiE) is a metric used to determine the energy efficiency of a data center by measuring what percentage of total facility power is consumed by IT equipment. https://cdn.ttgtmedia.com/visuals/digdeeper/5.jpg https://www.techtarget.com/searchdatacenter/definition/data-center-infrastructure-efficiency-DCIE Tue, 14 Oct 2025 17:15:00 GMT <h3>What is a unified computing system (UCS)?</h3> <p>A unified computing system (UCS) is a&nbsp;<a href="https://www.techtarget.com/searchdatacenter/definition/converged-infrastructure">converged</a>&nbsp;data center architecture that integrates computing, networking and storage resources to&nbsp;increase efficiency and enable centralized management. Modern UCS solutions are AI-ready infrastructure designed to support everything from traditional applications to AI model training and edge inferencing.</p> <p>UCS products are designed and configured to work together effectively. The goal of a UCS product line is to simplify the number of devices that need to be connected, configured, cooled and secured, enabling administrators to manage everything through a single graphical interface.</p> <p>The term <i>unified computing system</i> is often associated with Cisco. Cisco UCS products support traditional and <a href="https://www.techtarget.com/searchitoperations/definition/virtualization">virtualized</a>&nbsp;environments. Cisco UCS Manager is for on-premises deployments, and Cisco Intersight is for cloud-based management across the entire infrastructure lifecycle.</p> <p>Similar offerings to Cisco UCS include Hewlett Packard Enterprise Synergy and Dell VxRail.</p> A unified computing system (UCS) is a converged data center architecture that integrates computing, networking and storage resources to increase efficiency and enable centralized management. https://cdn.ttgtmedia.com/visuals/digdeeper/5.jpg https://www.techtarget.com/searchdatacenter/definition/unified-computing-system-UCS Tue, 14 Oct 2025 16:15:00 GMT <section class="section main-article-chapter" data-menu-title="What is statistical mean, median, mode and range?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What is statistical mean, median, mode and range?</h2> <p>The terms mean, median, mode and range describe properties of statistical distributions. In statistics, a distribution is the set of all possible values for terms that represent defined events. The value of a term, when expressed as a variable, is called a random variable.</p> <p>There are two major types of statistical distributions. The first type contains discrete random variables, which means every term has a precise, isolated numerical value. The second major type of distribution contains a continuous random variable. A continuous random variable is a random variable where the data can take infinitely many values. When a term can acquire any value within an unbroken interval or span, it is called a probability density function.</p> <p>IT professionals and <a href="https://www.techtarget.com/searchenterpriseai/definition/data-scientist">data scientists</a> need to understand the definition of mean, median, mode and range to&nbsp;<a href="https://www.techtarget.com/searchdatacenter/definition/data-center-capacity-planning">plan capacity and balance load</a>, manage systems, perform maintenance and troubleshoot issues.</p> </section> <section class="section main-article-chapter" data-menu-title="How are mean, median, mode and range used in the data center?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How are mean, median, mode and range used in the data center?</h2> <p>Many relevant tasks require the administrator to calculate mean, median, mode or range, or often some combination, to show a statistically significant quantity, trend or deviation from the norm.</p> <p>Finding the mean, median, mode and range is only the beginning of <a href="https://www.techtarget.com/searchdatacenter/definition/data-center-infrastructure-management-DCIM">data center management</a>. The administrator must then apply this information to investigate root causes of a problem, accurately forecast future needs or set acceptable working parameters for IT systems.</p> <p>When working with a large data set, it can be useful to represent the entire data set with a single value that describes the "middle" or "average" value of the entire set. In statistics, that single value is called the central tendency -- mean, median and mode are ways to describe it.</p> </section> <section class="section main-article-chapter" data-menu-title="Mean"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Mean</h2> <p>To calculate the mean, add the values in the data set and divide the sum by the number of values.</p> <p>The most common expression for the mean of a statistical distribution with a discrete random variable is the mathematical average of all the terms.</p> <p>The mean of a statistical distribution with a continuous random variable, also called the expected value, is obtained by integrating the variable's product with the distribution's probability. The lowercase Greek letter mu (µ) denotes the expected value.</p> </section> <section class="section main-article-chapter" data-menu-title="Median"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Median</h2> <p>To find the median, list the values of the data set in numerical order and identify which value appears in the middle of the list.</p> <p>The median of a distribution with a discrete random variable depends on whether the number of terms in the distribution is even or odd. If the number of terms is odd, then the median is the value of the term in the middle. This is the value such that the number of terms having values greater than or equal to it is the same as having values less than or equal to it.</p> <p>If the number of terms is even, then the median is the average of the two terms in the middle, such that the number of terms having values greater than or equal to it is the same as the number of terms having values less than or equal to it. For example, in a distribution with an even number of terms, with the middle terms being 5 and 7, the median value would be 6.</p> <p>The median of a distribution with a continuous random variable is the value m such that the probability is at least 1/2 (50%) that a randomly chosen point on the function will be less than or equal to m. The probability is also at least 1/2 that a randomly chosen point on the function will be greater than or equal to m.</p> </section> <section class="section main-article-chapter" data-menu-title="Mode"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Mode</h2> <p>To find the mode, identify which value in the data set occurs most often.</p> <p>The mode of a distribution with a discrete random variable is the value of the term that occurs the most often. It is not uncommon for a distribution with a discrete random variable to have more than one mode, especially if there are not many terms. This happens when two or more terms occur with equal frequency and more often than others.</p> <p>A distribution with two modes is called bimodal. A distribution with three modes is called trimodal. The mode of a distribution with a continuous random variable is the maximum value of the function.</p> </section> <section class="section main-article-chapter" data-menu-title="Range"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Range</h2> <p>To find the range, subtract the largest value from the smallest in the data set.</p> <p>Range describes how well the central tendency represents the data. If the range is large, the central tendency is not as representative of the data as it would be if the range were small.</p> <p>The range of a distribution with a discrete random variable is the difference between the maximum and minimum values. For a distribution with a continuous random variable, the range is the difference between the two extreme points on the distribution curve, where the value of the function falls to zero. For any value outside the range of a distribution, the value of the function is equal to 0.</p> </section> <section class="section main-article-chapter" data-menu-title="Using mean to determine power usage"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Using mean to determine power usage</h2> <p>To calculate the mean, add all the numbers in a set and then divide the sum by the total count of numbers. For example, in a data center rack, five&nbsp;<a href="https://www.techtarget.com/whatis/definition/server">servers</a>&nbsp;consume 100 watts, 98 watts, 105 watts, 90 watts and 102 watts of power, respectively. The mean power use of that rack is calculated as (100 + 98 + 105 + 90 + 102 W)/5 servers = a calculated mean of 99 W per server. Intelligent&nbsp;<a href="https://www.techtarget.com/searchdatacenter/definition/power-distribution-unit-PDU">power distribution units</a>&nbsp;report the mean power utilization of the rack to systems management software.</p> <div class="youtube-iframe-container"> <iframe id="ytplayer-0" src="https://www.youtube.com/embed/ajo1noUQk50?autoplay=0&amp;modestbranding=1&amp;rel=0&amp;widget_referrer=null&amp;enablejsapi=1&amp;origin=https://www.techtarget.com" type="text/html" height="360" width="640" frameborder="0"></iframe> </div> </section> <section class="section main-article-chapter" data-menu-title="Using median to plan capacity"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Using median to plan capacity</h2> <p>In the data center, means and medians are often tracked to spot trends that inform capacity planning or power cost predictions. The statistical median is the middle number in a sequence of numbers. To find the median, organize each number in order by size; the number in the middle is the median.</p> <p>For the five servers in the rack, arrange the power consumption figures from lowest to highest: 90 W, 98 W, 100 W, 102 W and 105 W. The median power consumption of the rack is 100 W. If there is an even set of numbers, average the two middle numbers. For example, if the rack had a sixth server that used 110 W, the new number set would be 90 W, 98 W, 100 W, 102 W, 105 W and 110 W. Find the median by averaging the two middle numbers: (100 + 102)/2 = 101 W.</p> </section> <section class="section main-article-chapter" data-menu-title="Using mode to identify a baseline"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Using mode to identify a baseline</h2> <p>Mode helps identify the most common or frequent occurrence of a characteristic. Within larger sets of numbers, there can be two modes (bimodal), three modes (trimodal) or more modes.</p> <p>For the server power consumption examples above, there is no mode because each element is different. But suppose the administrator measured the power consumption of an entire network operations center (<a href="https://www.techtarget.com/searchnetworking/definition/network-operations-center">NOC</a>) and the set of numbers is 90 W, 104 W, 98 W, 98 W, 105 W, 92 W, 102 W, 100 W, 110 W, 98 W, 210 W and 115 W. The mode is 98 W since that power consumption measurement occurs most often amongst the 12 servers.</p> </section> <section class="section main-article-chapter" data-menu-title="Using range to identify outliers"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Using range to identify outliers</h2> <p>The range is the difference between the highest and lowest values within a set of numbers. To calculate the range, subtract the smallest number from the largest number in the set. If a six-server rack includes 90 W, 98 W, 100 W, 102 W, 105 W and 110 W, the power consumption range is 110 W - 90 W = 20 W.</p> <p>Range shows how much the numbers in a set vary. Many IT systems operate within an acceptable range; a value more than that range&nbsp;<a href="https://www.techtarget.com/searchdatacenter/feature/Data-center-monitoring-best-practices-simplify-software-choice">might trigger a warning</a>&nbsp;or alarm to IT staff. To find the variance in a data set, subtract each number from the mean, and then square the result. Find the average of these squared differences, and that is the variance in the group. In our original group of five servers, the mean was 99. The 100 W-server varies from the mean by 1 W, the 105 W-server by 6 W, and so on. The squares of each difference equal 1, 1, 36, 81 and 9. So to calculate the variance, add 1 + 1 + 36 + 81 + 9 and divide by 5. The variance is 25.6. Standard deviation denotes how far apart all the numbers are in a set. The standard deviation is calculated by finding the square root of the variance. In this example, the standard deviation is 5.1.</p> <p>Interquartile range, the middle fifty or midspread of a set of numbers, removes the outliers -- the highest and lowest numbers in a set. If there is a large set of numbers, divide them evenly into lower and higher numbers. Then find the median of each of these groups. Find the interquartile range by subtracting the lower median from the higher median. If a rack of six servers' power wattage is arranged from lowest to highest: 90, 98, 100, 102, 105, 110, divide this set into low numbers (90, 98, 100) and high numbers (102, 105, 110). Find the median for each: 98 and 105. Subtract the lower median from the higher median: 105 watts - 98 W = 7 W, which is the interquartile range of these servers.</p> <p><b>Editor's note:</b>&nbsp;<i>This article was revised in 2025 by Informa TechTarget editors to improve the reader experience.</i></p> </section> The terms mean, median, mode, and range describe properties of statistical distributions. https://cdn.ttgtmedia.com/visuals/digdeeper/3.jpg https://www.techtarget.com/searchdatacenter/definition/statistical-mean-median-mode-and-range Thu, 09 Oct 2025 11:15:00 GMT <p>A graphics processing unit (GPU) is a computer <a href="https://www.techtarget.com/whatis/definition/microchip">chip</a> that renders graphics and images by performing rapid mathematical calculations. GPUs are used for both professional and personal computing. Originally, GPUs were responsible for the rendering of 2D and <a href="https://www.techtarget.com/whatis/definition/3-D-three-dimensions-or-three-dimensional">3D</a> images, animations and video, but now they have a wider use range.</p> <p>Like a central processing unit (<a href="https://www.techtarget.com/whatis/definition/processor">CPU</a>), a GPU is also a chip component in computing devices. One important difference is that the GPU is specifically designed to handle and accelerate graphics workloads and display graphics content on a device such as a PC or <a href="https://www.techtarget.com/searchmobilecomputing/definition/smartphone">smartphone</a>.</p> <p>An electronic device with an <a href="https://www.techtarget.com/whatis/definition/embedded-device">embedded</a> or discrete GPU can smoothly render 3D graphics and video content, making it suitable for gaming and other visual applications. Over time, technological improvements have resulted in more flexible and programmable GPUs that can be used for many more applications and workloads other than gaming. GPUs are now used for creative content production, video editing, high-performance computing (<a href="https://www.techtarget.com/searchdatacenter/definition/high-performance-computing-HPC">HPC</a>) and artificial intelligence (<a href="https://www.techtarget.com/searchenterpriseai/definition/AI-Artificial-Intelligence">AI</a>).</p> <section class="section main-article-chapter" data-menu-title="What does a GPU do?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What does a GPU do?</h2> <p>In the early days of computing, the CPU performed the calculations required for graphics applications, such as the rendering of 2D and 3D images, animations and video. As more graphics-intensive applications were developed, however, their demands put a strain on the CPU and decreased the computer's overall performance.</p> <p>Today, a GPU is a specialized computing system that performs these tasks:</p> <ul class="default-list"> <li>Handles graphics-related tasks in lieu of the CPU.</li> <li>Performs graphics calculations very quickly.</li> <li>Uses <a href="https://www.techtarget.com/searchdatacenter/definition/parallel-processing">parallel processing</a> to increase processing performance.</li> <li>Delivers graphic content to the computer display.</li> <li>Frees up the CPU to handle all other processing tasks.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="How does a GPU work?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How does a GPU work?</h2> <p>GPUs work by using parallel processing, where multiple processors handle separate parts of a single task. A GPU will also have its own <a href="https://www.techtarget.com/searchstorage/definition/RAM-random-access-memory">RAM</a> to store the data it is processing. This RAM is designed specifically to hold the large amounts of information coming into the GPU for highly intensive graphics use cases.</p> <p>For graphics applications, the CPU sends instructions to the GPU for drawing the graphics content on a screen. The GPU executes the instructions in parallel and at high speeds to display the content on the device -- a process known as the graphics or rendering <a href="https://www.techtarget.com/whatis/definition/pipelining">pipeline</a>.</p> <div class="youtube-iframe-container"> <iframe id="ytplayer-0" src="https://www.youtube.com/embed/_DNxgnINNAY?autoplay=0&amp;modestbranding=1&amp;rel=0&amp;widget_referrer=null&amp;enablejsapi=1&amp;origin=https://www.techtarget.com" type="text/html" height="360" width="640" frameborder="0"></iframe> </div> </section> <section class="section main-article-chapter" data-menu-title="Are GPUs and graphics cards the same?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Are GPUs and graphics cards the same?</h2> <p>GPU and graphics card are two terms that are sometimes used interchangeably. However, there are some important distinctions between the two. The main difference is that the GPU is a specific unit within a graphics card, among other components. The GPU is what performs the image and graphics processing. The graphics card presents images to the display unit.</p> <h3>GPU use cases: What GPUs are used for today</h3> <p>GPUs are widely used for PC gaming, allowing for smooth, high-quality graphics rendering. Modern GPUs are also adapted to a wider variety of tasks than they were originally designed for, partially because they are more programmable than they were in the past. That's why GPUs are now also used to <a href="https://www.techtarget.com/searchenterpriseai/feature/CPUs-vs-GPUs-for-AI-workloads">accelerate AI workloads</a> and for machine learning (<a href="https://www.techtarget.com/searchenterpriseai/definition/machine-learning-ML">ML</a>).</p> <p>Some of the most popular applications of GPUs include:</p> <ul class="default-list"> <li><b>Rendering real-time 2D and 3D graphics</b>. These devices perform complicated mathematical calculations using <a href="https://www.techtarget.com/whatis/definition/algorithm">algorithms</a> that convert moving bits into <a href="https://www.techtarget.com/whatis/definition/pixel">pixels</a> on displays.</li> <li><b>Video editing and video content creation</b>. When used for video content creation and editing, GPUs deliver high-resolution images and videos, enhanced video effects using AI, and provide sufficient RAM to handle the computational requirements of video content.</li> <li><b>Video game graphics.</b> Video game GPUs perform the same kinds of complex calculations using parallel processing as for other graphics tasks, and deliver images and videos by managing tasks such as shading, lighting and texturing, each of which is needed for realistic game visuals.</li> <li><b>Accelerating AI/ML applications.</b> GPUs have the processing power needed to support the massive computations and algorithms used in AI and ML applications, such as image recognition and <a href="https://www.techtarget.com/searchenterpriseai/definition/face-detection">facial detection</a> and <a href="https://www.techtarget.com/searchenterpriseai/definition/facial-recognition">recognition</a>.</li> <li><b>Training deep learning neural networks</b>. GPUs have thousands of smaller core memory units that work in parallel to provide the matrix processing needed for <a href="https://www.techtarget.com/searchenterpriseai/definition/deep-learning-deep-neural-network">deep learning</a> training and <a href="https://www.techtarget.com/searchenterpriseai/definition/neural-network">neural networks</a>.</li> <li><b>Cryptomining. </b>GPUs support cryptocurrency mining, especially for coins that employ Proof of Work (<a href="https://www.techtarget.com/whatis/feature/Proof-of-work-vs-proof-of-stake-Whats-the-difference">PoW</a>) mechanisms for establishing consensus; they also use parallel processing to perform multiple <a href="https://www.techtarget.com/searchdatamanagement/definition/hashing">hash</a> functions.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="Types of GPUs"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Types of GPUs</h2> <p>Generally, there are three types of GPUs:</p> <ul class="default-list"> <li><b>Integrated GPU. A</b>n integrated GPU is built into the computer's <a href="https://www.techtarget.com/whatis/definition/motherboard">motherboard</a> or could be integrated with the CPU. Systems with integrated GPUs are generally small and light because less space is required to incorporate the GPU. The use of an internal GPU also reduces system power consumption. When buying a system with an internal GPU, care must be taken as the internal GPU may not be upgradeable. If buying a gaming computer, the GPU is an essential part of the device and should be carefully researched.</li> <li><b>Discrete GPU.</b> This is a physically separate GPU and is mounted separately from the CPU and other peripherals on the motherboard. While such devices typically offer far more processing power than an integrated GPU, they also increase power usage and generate additional heat, which may necessitate additional cooling. Discrete GPUs are used for resource-intensive, high-performance applications such as 3D games.</li> <li><b>Hybrid GPU.</b> A hybrid GPU arrangement can take different forms, such as combining an integrated GPU with a discrete GPU in the same system, with one used for normal graphics and the discrete unit for high-performance requirements. Another approach is to combine discrete graphics performance with specialized accelerators, such as for AI applications.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="What is a cloud GPU?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What is a cloud GPU?</h2> <p>Taking advantage of the convenience, <a href="https://www.techtarget.com/searchdatacenter/definition/scalability">scalability</a> and costs of <a href="https://www.techtarget.com/searchcloudcomputing/definition/cloud-computing">cloud computing</a> services, users can obtain the GPU functionality they need by obtaining it from a cloud vendor. All three major cloud services, AWS, Microsoft Azure and Google Cloud, offer virtual machines (<a href="https://www.techtarget.com/searchitoperations/definition/virtual-machine-VM">VMs</a>) with built-in GPU speed and functionality. Cloud GPU services are typically used in compute-intensive situations involving AI, video game development and <a href="https://www.techtarget.com/whatis/definition/Bitcoin-mining">bitcoin mining</a>. As with any activity involving cloud services, cybersecurity must be carefully managed.</p> </section> <section class="section main-article-chapter" data-menu-title="GPU vs. CPU"> <h2 class="section-title"><i class="icon" data-icon="1"></i>GPU vs. CPU</h2> <p>A GPU might be found integrated with a CPU on the same electronic circuit, on a graphics card or in the motherboard of a PC or <a href="https://www.techtarget.com/whatis/definition/server">server</a>. GPUs and CPUs are fairly similar in construction. However, CPUs are used to respond to and process the basic instructions that drive a computer, while GPUs are designed specifically to quickly render high-resolution images and video.</p> <p>Essentially, CPUs are responsible for interpreting most of a computer's commands, while GPUs perform more complex mathematical and geometric calculations to focus on graphics rendering and other applications that require intensive calculations.</p> <p>Both processors have different numbers of cores and transistors. The core can be thought of as the processor within the processor. Each core can process its own tasks, or <a href="https://www.techtarget.com/whatis/definition/thread">threads</a>. A CPU uses fewer cores and performs tasks sequentially. A GPU, in contrast, might have hundreds or thousands of cores, which allow for parallel processing and lightning-fast graphics output.</p> <p>A single-core CPU usually lacks the capability for parallel processing, but <a href="https://www.techtarget.com/searchdatacenter/definition/multi-core-processor">multicore processors</a> can perform calculations in parallel by combining more than one CPU onto the same chip. GPUs can also contain more transistors than a CPU.</p> <p>In addition, a CPU has a higher <a href="https://www.techtarget.com/whatis/definition/clock-speed">clock speed</a>, meaning it can perform an individual calculation faster than a GPU, so it is often better equipped to handle basic computing tasks.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/enterprise_ai-cpu_vs_gpu_chips.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/enterprise_ai-cpu_vs_gpu_chips_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/enterprise_ai-cpu_vs_gpu_chips_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/enterprise_ai-cpu_vs_gpu_chips.png 1280w" alt="CPU vs. GPU chips" height="370" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>While their architectures and technologies are similar, CPUs and GPUs perform discrete functions, with GPUs focused on calculation-intensive computations. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <p><i>Data centers use three processing units: CPU, GPU, and DPU. Each has unique functions that can be combined to enhance data center performance. Explore </i><a href="https://www.techtarget.com/searchdatacenter/tip/How-do-CPU-GPU-and-DPU-differ-from-one-another"><i>how CPU vs. GPU vs. DPU differ from one another</i></a><i>.</i></p> </section> A graphics processing unit (GPU) is a computer chip that renders graphics and images by performing rapid mathematical calculations. https://cdn.ttgtmedia.com/visuals/digdeeper/1.jpg https://www.techtarget.com/searchvirtualdesktop/definition/GPU-graphics-processing-unit Fri, 05 Sep 2025 13:00:00 GMT <p>Seoul-based chip startup FuriosaAI is looking to threaten Nvidia's data center AI chip dominance and has scored its first major contract -- powering LG's Exaone large language model.</p> <p>LG AI Research will offer FuriosaAI's RNGD chip-powered servers to enterprise customers across electronics, finance, telecommunications and biotechnology for <a href="https://www.techtarget.com/whatis/definition/large-language-model-LLM">LLMs</a>.</p> <p>FuriosaAI's <a target="_blank" href="https://developer.furiosa.ai/latest/en/overview/rngd.html" rel="noopener">RNGD chip</a>, a neural processing unit, is designed to support LLMs and other deep learning models on the inference side, rather than training. <a href="https://www.techtarget.com/searchenterpriseai/tip/AI-inference-vs-training-Key-differences-and-tradeoffs">Inference uses trained models</a> to make predictions.</p> <p>RNGD is based on tensor contraction processor architecture and is manufactured using Taiwan Semiconductor Manufacturing Co.'s 5 nm process node. The company claims that RNGD achieves 2.25 times better per-watt performance for inference over GPUs, which have become the most popular processors for AI workloads.</p> <p>"I think FuriosaAI is smart to go after the inference market as its primary target ... It is far more open and lucrative than the training market," said Matt Kimball, vice president and principal analyst at Moor Insights &amp; Strategy, in an interview.</p> <section class="section main-article-chapter" data-menu-title="Nvidia's hold on data center"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Nvidia's hold on data center</h2> <p>Toppling GPU juggernaut Nvidia is easier said than done. Nvidia enjoyed a 98% data center GPU <a target="_blank" href="https://siliconanalysts.com/nvidia-ai-accelerator-market-outlook-2023-2027" rel="noopener">market share in 2023</a>, according to Silicon Analysts. Competitors claiming performance wins over Nvidia's hardware still have a mountain to climb -- Nvidia's popular CUDA software helps the company maintain its comfortable grip on AI workloads. GPU challenges from AMD and Intel have done little to loosen that grip.</p> <p>"Training is dominated by Nvidia GPUs and CUDA, with AMD showing potential to eat some of that share on the enterprise side," Kimball said. "Does [FuriosaAI's RNGD] pose a threat to Nvidia? I don't think so. ... RNGD seems to have scored in its niche area."</p> <p>Nvidia leads in the inference market with its A100, H100, L4, L40 and Blackwell GPUs. But the inference side is more diverse, with competition from Google using custom-built TPU v4i chips, Amazon with in-house Inferentia chips, Intel CPUs, AMD, Qualcomm and several startup chipmakers.</p> </section> <section class="section main-article-chapter" data-menu-title="Finding a niche in AI inference"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Finding a niche in AI inference</h2> <p>Inference is gaining momentum and is expected to be a larger market than training in the coming years. The global AI inference market was estimated at $106 billion in 2025 and projected to grow to $255 billion by 2030, according to a MarketsandMarkets <a target="_blank" href="https://www.marketsandmarkets.com/Market-Reports/ai-inference-market-189921964.html" rel="noopener">report</a>.</p> <p>Kimball said FuriosaAI's technology is aimed at a very specific low-latency inference, and that Nvidia's GPUs will still be valuable on the inference side, especially for extremely large models.</p> <p>"RNGD has an architecture that lends itself to high-query-per-second and lower-precision kind of inference. ... Think recommendation engines, customer service bots, programmatic advertising, etc.," Kimball said.</p> <p>Other big tech players clearly see potential in the growing inference market. In March, FuriosaAI turned down Meta's $800 million bid to acquire the company.</p> <p>"I believe the inference market is wide open. Partly because of its size, and partly because of the diversity of use cases and deployment scenarios," Kimball said.</p> <p><i>Shane Snider, a veteran journalist with more than 20 years of experience, covers IT infrastructure at Informa TechTarget.</i></p> </section> The South Korean semiconductor startup scored LG as its first major customer, as companies compete to take advantage of the AI boom. https://cdn.ttgtmedia.com/rms/onlineimages/maze_g676210320.jpg https://www.techtarget.com/searchdatacenter/news/366627769/FuriosaAI-to-fuel-LG-Exaone-LLM-Is-it-a-challenge-to-Nvidia Tue, 22 Jul 2025 16:37:00 GMT <p>A power distribution unit (PDU) is a device for controlling <a href="https://www.techtarget.com/searchdatacenter/definition/data-center">data center</a> electrical power. The most basic PDUs are large power strips without surge protection. They are designed to provide standard electrical outlets for data center equipment and have no monitoring or remote access capabilities. More advanced PDUs provide real-time monitoring and remote access capabilities.</p> <p>Power distribution units are designed to forward alternating current (AC) from a power source to multiple devices in the data center. PDUs can be portable, mounted on the data center floor or installed directly on a data center <a href="https://www.techtarget.com/whatis/definition/rack">rack</a>. Data center devices plug into outlets on the PDU to receive their electrical power.</p> <p>PDUs are used in environments where ordinary power strips and <a href="https://www.techtarget.com/whatis/definition/surge-suppressor-surge-protector">surge protectors</a> cannot meet electrical power requirements safely and/or reliably. Organizations with complex data centers can use intelligent PDUs to manage power to specific outlets on the PDU remotely and improve <a href="https://www.techtarget.com/whatis/definition/uptime-and-downtime">uptime</a> by identifying and resolving power issues before they cause equipment failures.</p> <p>According to a report published by Allied Market Research in 2025, the <a target="_blank" href="https://www.openpr.com/news/4045950/power-distribution-unit-market-set-to-surge-with-industrial" rel="noopener">market for PDUs</a> is expected to reach $7.9 billion by 2030. Industry experts often attribute the continued growth of this market to a growing reliance on <a href="https://www.techtarget.com/searchcloudcomputing/definition/cloud-computing">cloud computing</a> and the rapid expansion of data centers to meet the needs of <a href="https://www.techtarget.com/searchcio/definition/digital-transformation">digital transformation</a> initiatives.</p> <section class="section main-article-chapter" data-menu-title="What do PDUs do?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What do PDUs do?</h2> <p>PDUs do essentially the same job as power strips: They supply power to multiple items at once. Data centers and other industrial environments cannot plug all their devices into one power supply -- there are just too many of them. Instead, PDUs distribute power to multiple devices, including <a href="https://www.techtarget.com/searchdatacenter/answer/How-do-I-estimate-server-power-consumption-per-rack">servers</a>, routers, <a href="https://www.techtarget.com/searchnetworking/definition/switch">switches</a> and telecommunications equipment. One PDU can provide power for up to 48 devices simultaneously, depending on the form factor and use case.</p> <p>In most data center environments, a single PDU is installed on each rack to distribute power to all the equipment housed within that rack. This setup simplifies power management and helps administrators organize power delivery on a per-rack basis.</p> <p>It should be noted, however, that racks containing mission-critical data center components are often equipped with more than one PDU, each of which connects to a separate power source. This allows <a href="https://www.techtarget.com/searchstorage/definition/failover">failover</a> and continued operation if one power source or PDU fails.</p> <p>Racks that contain <a href="https://www.techtarget.com/searchdatacenter/definition/blade-server">blade servers</a> or graphics processing unit (<a href="https://www.techtarget.com/searchvirtualdesktop/definition/GPU-graphics-processing-unit">GPU</a>)-intensive systems might also require multiple PDUs to deliver enough current safely. Blade servers and GPU workloads draw more <a href="https://www.techtarget.com/whatis/definition/current">current</a> per unit of rack space compared to standard servers, which can lead to load imbalances or even overload conditions if only one PDU is used.</p> </section> <section class="section main-article-chapter" data-menu-title="How do PDUs work?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How do PDUs work?</h2> <p>While all PDUs serve the basic function of delivering power to multiple devices, they vary significantly in terms of monitoring capabilities and control features.</p> <p><b>Basic PDUs.</b> These units work the same way as a heavy-duty power strip. They distribute power from a single input source to multiple output outlets without any monitoring or remote control. Basic PDU units are commonly used in small businesses that prioritize simplicity and cost-effectiveness.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/cyberpower_pdu_15b2f10r_f.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/cyberpower_pdu_15b2f10r_f_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/cyberpower_pdu_15b2f10r_f_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/cyberpower_pdu_15b2f10r_f.png 1280w" alt="Image of CyberPower PDU that has two outlets and a power button on the front." height="147" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>The CyberPower PDU15B2F10R is a basic rackmount power distribution unit (PDU). It has 12 outlets -- two in the front, 10 in the rear -- and provides 120-volt, 15-amp output with unfiltered electrical pass-through. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <p><b>Metered PDUs. </b>These work like basic PDUs, but they have a digital display that shows real-time data about electrical power consumption. Metered PDUs are helpful for load balancing and avoiding overcapacity issues, especially when metering is done at the outlet level.</p> <p><b>Monitored PDUs.</b> Sometimes called intelligent PDUs, these units give administrators real-time visibility into power usage, but they don't allow admins to remotely control individual outlets.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/cyberpower_pdu_41003_f.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/cyberpower_pdu_41003_f_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/cyberpower_pdu_41003_f_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/cyberpower_pdu_41003_f.png 1280w" alt="Image shows a CyberPower monitored PDU that has 16 output receptacles on the front. " height="157" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>The CyberPower PDU41003 is a monitored PDU that provides 16 output receptacles. This PDU includes a built-in web-based management interface that can be accessed using the PDU's IP address in a web browser. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <p><b>Switched PDUs.</b> These are like metered PDUs that can be controlled remotely. In addition to providing metered power data, this type of PDU lets administrators remotely power devices on or off at individual outlets. This is especially useful for rebooting unresponsive equipment remotely.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/dps_telecom_ab6_front_2022.jpg"> <img data-src="https://www.techtarget.com/rms/onlineimages/dps_telecom_ab6_front_2022_mobile.jpg" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/dps_telecom_ab6_front_2022_mobile.jpg 960w,https://www.techtarget.com/rms/onlineimages/dps_telecom_ab6_front_2022.jpg 1280w" alt="This image shows a switch power distribution unit that was designed for telecom data centers. " height="96" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>The DPS Telecom's Remote Power Switch AB6 has a Smart Fuse Panel with status lights. Each fuse in the panel is a safety device that melts if the current flowing through it exceeds a certain threshold. Unlike most switched PDUs used in data centers, the AB6 is optimized for DC power distributions and telecom voltage requirements. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <p><b>Automatic transfer switch (ATS) PDUs.</b> ATS units provide redundancy for single-corded devices. This type of PDU has dual power inputs and can automatically switch from the primary power source to the secondary source if a failure is detected.</p> <p><b>Dual-circuit PDUs.</b> These have two power cords, each of which feeds a separate set of outlets. This enables devices that already have dual power supplies to plug into both circuits for redundancy.</p> <p><b>High-density PDUs.</b> These units are built for data centers that run <a href="https://www.techtarget.com/searchdatacenter/definition/high-performance-computing-HPC">high-performance computing</a> workloads or have dense server configurations. This type of PDU can handle very high power loads and typically offers different outlet configurations to support a wide range of devices.</p> <p><b>Specialized PDUs.</b> These are hybrid industrial PDUs that can often deliver AC and direct current (DC) distributions simultaneously. DC power is used in specialized environments, such as telecom facilities or <a href="https://www.techtarget.com/searchcloudcomputing/definition/hyperscale-cloud">hyperscale cloud</a> data centers that have DC power architectures.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/dps_telecom_cabinet_pdu_tall.jpg"> <img data-src="https://www.techtarget.com/rms/onlineimages/dps_telecom_cabinet_pdu_tall_mobile.jpg" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/dps_telecom_cabinet_pdu_tall_mobile.jpg 960w,https://www.techtarget.com/rms/onlineimages/dps_telecom_cabinet_pdu_tall.jpg 1280w" alt="This image shows a DPS Telecom PDU that can handle both AC and DC power distribution. " height="373" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>DPS Telecom's AC+DC DIN PDU is designed to mount on a standard DIN rail in cabinets. This PDU features a single AC input that supplies power to six AC outputs. Its six 24 VDC outputs make it suitable for applications that require both AC and DC power distribution. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> </section> <section class="section main-article-chapter" data-menu-title="What are the different types of PDU form factors?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What are the different types of PDU form factors?</h2> <p>Power distribution units come in several <a href="https://www.techtarget.com/whatis/definition/form-factor">form factors</a> to suit different environments and power needs. Each type of PDU plays a specific role depending on an organization's power management needs, scalability goals and operational priorities.</p> <p><b>Rackmount PDUs.</b> These are used in data centers to deliver power to rack-mounted data center equipment. Advanced intelligent models can monitor and control power at the outlet level and assist with <a href="https://www.techtarget.com/searchnetworking/definition/load-balancing">load balancing</a>.</p> <p><b>Floor-standing PDUs.</b> These serve as intermediate distribution points between upstream power from an <a href="https://www.techtarget.com/searchdatacenter/definition/uninterruptible-power-supply">uninterruptible power supply</a> or generator and rack-mounted PDUs. Floor-standing PDUs might also be referred to as <i>remote power panels</i>.</p> <p><b>Power distribution cabinets.</b> These floor-standing PDUs include circuit breakers and electrical monitoring panels. This type of PDU is typically used in high-density, enterprise-class data centers to manage large volumes of power safely and efficiently.</p> <p><b>Portable PDUs.</b> Portable units are not mounted to racks and are intended for light-duty power distribution. They resemble industrial-grade power strips and are used in settings where flexible, mobile power distribution is required.</p> </section> <section class="section main-article-chapter" data-menu-title="Basic PDUs vs. intelligent PDUs"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Basic PDUs vs. intelligent PDUs</h2> <p>Basic PDUs are essentially high-capacity power strips designed for use in industrial or IT environments. They distribute power from a single source to multiple outlets and enable multiple devices to be powered from a central point. Basic PDUs are typically used in environments where cost efficiency is important and the requirements for power distribution are straightforward.</p> <p>In contrast, intelligent PDUs, which might also be referred to as <i>smart</i> or <i>managed PDUs</i>, often have real-time power monitoring features that let IT administrators remotely monitor <a href="https://www.techtarget.com/whatis/definition/voltage">voltage</a>, current and power usage through a web-based interface or <a href="https://www.techtarget.com/searchnetworking/definition/SNMP">Simple Network Management Protocol</a> tools. If the intelligent PDU supports remote power cycling, administrators can also reboot individual devices that connect to the PDU without having to be physically present.</p> <p>Some high-end intelligent PDUs can share monitoring data with the organization's data center infrastructure management (<a href="https://www.techtarget.com/searchdatacenter/definition/data-center-infrastructure-management-DCIM">DCIM</a>) system. This capability is especially valuable in <a href="https://www.techtarget.com/searchdatacenter/definition/lights-out-management">lights-out</a> data centers and <a href="https://www.techtarget.com/searchcloudcomputing/definition/What-is-distributed-cloud-and-how-does-it-work">distributed cloud</a> data center environments where uptime, energy efficiency and remote management capabilities are important.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/basic_vs_intelligent_power_distribution_units-f.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/basic_vs_intelligent_power_distribution_units-f_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/basic_vs_intelligent_power_distribution_units-f_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/basic_vs_intelligent_power_distribution_units-f.png 1280w" alt="This comparison chart shows the difference between basic PDUs and intelligent PDUs. " height="330" width="559"> <figcaption> <i class="icon pictures" data-icon="z"></i>Basic PDUs and intelligent PDUs both distribute electrical power to multiple devices, but intelligent PDUs offer advanced features such as real-time power monitoring, remote outlet control and integration with data center management systems. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <h3>Benefits of intelligent PDUs</h3> <p>Unlike basic PDUs that simply distribute electrical power, intelligent PDUs enable IT administrators to remotely monitor power consumption at the device or outlet level. The monitoring data can be used to identify which racks are using the most electricity and which racks are not being used to their full capacity. Many intelligent PDUs also support remote outlet control capabilities that let administrators turn off devices that are not in use. In a large data center, this can reduce both energy waste and operating costs.</p> <p>Intelligent PDUs also play an important role in calculating a data center's power usage effectiveness (<a href="https://www.techtarget.com/searchdatacenter/definition/power-usage-effectiveness-PUE">PUE</a>). PUE is a metric that quantifies how efficiently a data center uses electricity. It's calculated, as noted in the following equation, by dividing the total power entering the data center by the power used by the IT equipment:</p> <p style="text-align: center;">PUE = Total facility power / IT equipment power</p> </section> <section class="section main-article-chapter" data-menu-title="The impact of AI on PDUs"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The impact of AI on PDUs</h2> <p>Artificial intelligence has made it easier and more cost-effective to integrate and analyze the data collected by intelligent PDUs with other systems. For example, DCIM platforms and cloud-based analytics tools can combine PDU data with other infrastructure data to do the following:</p> <ul class="default-list"> <li>Analyze trends in power consumption.</li> <li>Optimize workload placement for energy efficiency.</li> <li>Predict failures or overload risks.</li> <li>Automate responses to power anomalies.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="Challenges of using PDUs"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Challenges of using PDUs</h2> <p>Power distribution units offer many advantages, but their use in data centers also comes with challenges that IT teams need to manage carefully.</p> <p>One of the biggest challenges is keeping the workflow for power planning and provisioning as simple as possible. As data center workloads grow and equipment densities increase, it becomes more difficult for administrators to predict how much power will be needed per rack and distribute power efficiently without overloading circuits or wasting capacity.</p> <p>Cable management and physical installations are another challenge. High-density PDUs, especially rack-mounted PDUs with lots of outlets, can create cable congestion. When too many cables are routed in close proximity, it can restrict the airflow needed to cool servers and other data center devices.</p> <p>Cost is also a factor. While basic PDUs are relatively inexpensive, intelligent PDUs that can be integrated with the organization's DCIM system can cost thousands of dollars more. Not all PDUs use the same communication protocols or interfaces, and organizations might need to invest in middleware or purchase <a target="_blank" href="https://www.raritan.com/landing/customized-rack-pdu" rel="noopener">customized PDUs</a> to enable seamless integration.</p> <p>Finally, it's important to understand that security is an increasing concern for intelligent PDUs. Like other network devices, intelligent PDUs can be vulnerable to unauthorized access or <a href="https://www.techtarget.com/searchsecurity/tip/6-common-types-of-cyber-attacks-and-how-to-prevent-them">cyberattacks</a> if they are not properly secured.</p> </section> <section class="section main-article-chapter" data-menu-title="PDU manufacturers"> <h2 class="section-title"><i class="icon" data-icon="1"></i>PDU manufacturers</h2> <p>Strong access controls and network segmentation can help protect PDUs, but a thoughtful approach to choosing the right vendor and type of PDU is also important. Numerous companies manufacture power distribution units. Here is a brief list of manufacturers:</p> <ul class="default-list"> <li><a target="_blank" href="https://www.apc.com/shop/us/en/tools/rack-pdu-selector" rel="noopener">APC by Schneider Electric</a>.</li> <li><a target="_blank" href="https://www.aten.com/us/en/products/power-distribution/rack-pdu/" rel="noopener">Aten</a>.</li> <li><a target="_blank" href="https://www.belden.com/products/racks-cabinets-cable-management/power-distribution-pdus#sort=%40catalogitemwebdisplaypriority%20ascending&amp;numberOfResults=25" rel="noopener">Belden</a>.</li> <li><a target="_blank" href="https://www.cyberpowersystems.com/products/pdus/" rel="noopener">CyberPower</a>.</li> <li><a target="_blank" href="https://www.deltapowersolutions.com/en/mcis/data-center-power-distribution-cabinet.php" rel="noopener">Delta Electronics</a>.</li> <li><a target="_blank" href="https://www.eaton.com/us/en-us/products/backup-power-ups-surge-it-power-distribution/power-distribution-for-it-equipment.html" rel="noopener">Eaton</a>.</li> <li><a target="_blank" href="https://www.enconnex.com/catalog/Power-Distribution-Units-PDUs-c133989019" rel="noopener">Enconnex</a>.</li> <li><a target="_blank" href="https://www.raritan.com/product-selector/pdu-selector" rel="noopener">Raritan</a>.</li> <li><a target="_blank" href="https://www.servertech.com/products?utm_source=google&amp;utm_medium=cpc&amp;utm_campaignID=701VS00000OR1iBYAT&amp;gad_source=1&amp;gad_campaignid=319125143&amp;gbraid=0AAAAAD_uW8MblyUU-vfLYdArzXs-VLlqR&amp;gclid=Cj0KCQjwgIXCBhDBARIsAELC9ZhS-jSV3ENFDkEbq_ax7JouQ6G0m06dHpeLaRV5VQgXN6s0N_p4V-AaAp7jEALw_wcB" rel="noopener">Server Technology</a>.</li> <li><a target="_blank" href="https://www.vertiv.com/en-in/products-catalog/critical-power/power-distribution/#/" rel="noopener">Vertiv</a>.</li> </ul> <p>When purchasing a PDU, it's important to understand that cost considerations should not be limited to the PDU's purchase price. Installation, maintenance requirements and warranties should all factor into <a href="https://www.techtarget.com/searchdatacenter/definition/TCO">total cost of ownership</a> and influence both vendor selection and model selection.</p> </section> <section class="section main-article-chapter" data-menu-title="Choosing the right PDU"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Choosing the right PDU</h2> <p>Before acquiring a new power distribution unit, an organization must decide on the type of PDU it needs. Technical specifications to look at and questions to ask vendors when making this decision include the following:</p> <ul class="default-list"> <li><b>Where will the PDU be installed?</b> It's important to determine whether the PDU will be mounted on a rack, placed on the floor or housed in a cabinet. Be sure to consider factors such as vertical or horizontal orientation, the amount of physical space the PDU takes up and airflow requirements.</li> <li><b>What type of input power is required?</b> Consider whether the PDU will only handle AC input or whether it will also need to handle DC input. Check to make sure the PDU under consideration is actually rated for voltages used in the data center.</li> <li><b>How much power does the equipment need?</b> Calculate the total power draw of the devices connected to the PDU to ensure the unit's maximum load rating will not be exceeded. This requires adding the wattage of all devices that will be connected to the PDU together and comparing the sum with the PDU's maximum load rating. To maintain a safe operating margin, the wattage sum should not exceed 80% of the PDU's rated capacity for continuous use.</li> <li><b>How many outlets are required?</b> Choose a PDU that has enough outlets to meet current needs, as well as extra capacity to accommodate future needs.</li> <li><b>What types of plugs and outlets are needed?</b> Make sure the PDU supports the plug configurations of all connected devices. Some PDUs offer mixed outlet types to provide flexibility.</li> <li><b>What level of redundancy is needed?</b> In mission-critical environments, consider purchasing additional PDUs to serve as failover units.</li> <li><b>Are advanced features required?</b> Intelligent PDUs have varying features and price points. Decide ahead of time whether specific features, such as the ability to integrate with the organization's DCIM, are necessary.</li> </ul> <div class="youtube-iframe-container"> <iframe id="ytplayer-0" src="https://www.youtube.com/embed/WpaM7z9TPuo?autoplay=0&amp;modestbranding=1&amp;rel=0&amp;widget_referrer=null&amp;enablejsapi=1&amp;origin=https://www.techtarget.com" type="text/html" height="360" width="640" frameborder="0"></iframe> </div> </section> <section class="section main-article-chapter" data-menu-title="Installing a PDU"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Installing a PDU</h2> <p>The specific steps involved in installing a power distribution unit will vary depending on the type and model. For a standard rackmount PDU, the general installation process typically includes the following steps:</p> <ol class="default-list"> <li>Determine whether the unit will be mounted vertically (0U) or horizontally (1U or 2U) in the rack. Align the appropriate mounting brackets and attach them securely to the PDU.</li> <li>Next, install the PDU by fastening the brackets to the rack rails. Once the PDU is mounted, connect its power cord to the appropriate power source. Be sure to route the power cord carefully to avoid impacting airflow or service paths. Consider using cable ties or straps to prevent strain on the connectors.</li> <li>After the power distribution unit is connected to power, it's time to connect servers and other IT devices to the PDU's outlets.</li> <li>Finally, turn on the PDU's power (if it has a power switch) and verify that each outlet is functioning correctly. Administrators for intelligent PDUs can use the PDU's <a href="https://www.techtarget.com/searchapparchitecture/definition/user-interface-UI">user interface</a> to verify outlet status and electrical power consumption.</li> </ol> </section> <section class="section main-article-chapter" data-menu-title="PDU vs. PSU: What is the difference?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>PDU vs. PSU: What is the difference?</h2> <p>The terms <i>PDU</i> and <i>PSU</i> (power supply unit) are often confused. Both types of units deal with power, but they have very different purposes. A PDU distributes electrical power to multiple devices, while a PSU is a device component that converts the power it receives into a usable format.</p> <p>In most data centers, AC is the standard form of electricity supplied by utility companies. Once the AC power reaches each device, however, it needs to be converted into DC power because that's what internal computer components use to operate. Essentially, the PDU delivers power to devices, and a PSU makes that power usable within each individual device.</p> <p><em>Learn why the future of the electrical grid is expected to be closely tied to <a href="https://www.techtarget.com/searchdatacenter/tip/How-data-centers-can-help-balance-the-electrical-grid">how data centers use electricity</a>.</em></p> </section> A power distribution unit (PDU) is a device for controlling data center electrical power. https://cdn.ttgtmedia.com/visuals/digdeeper/2.jpg https://www.techtarget.com/searchdatacenter/definition/power-distribution-unit-PDU Thu, 26 Jun 2025 00:00:00 GMT <p>Data center tiers are a system created by Uptime Institute LLC to describe the availability of infrastructure resources in a facility. A division of 451 Group, the Uptime Institute is the standard for this certification effort, which defines what is required to ensure that <a href="https://www.techtarget.com/searchdatacenter/definition/data-center">data centers</a> meet an organization's business demands. The tiering system came into place in the mid-1990s, making choosing a <a href="https://www.datacenterknowledge.com/ai-data-centers/afcom-ai-boom-fueling-data-center-construction-design-innovation">data center design and construction</a>, whether it be public or private, a lot easier when it comes to predicted <a href="https://www.techtarget.com/whatis/definition/uptime-and-downtime">uptime</a> and availability.</p> <p>There are four tiers, ranked by Roman numerals from Tier I -- the most basic data center design meeting the Institute's core requirements -- to Tier IV -- the most <a href="https://www.techtarget.com/searchdisasterrecovery/definition/fault-tolerant">fault-tolerant</a> and available data center design. Note that only Roman lettering is allowed; any data center advertising itself with Arabic numerals -- Tier 3, for example -- should be reviewed as it might not be certified by the Uptime Institute.</p> <p>As the tier levels increase, organizations must make bigger investments to meet certification criteria. Organizations should prepare a <a href="https://www.techtarget.com/searchsecurity/definition/risk-assessment">risk assessment</a> that balances business risk against cost and use this to decide which tier is right for them. The Uptime Institute has its own professional services group that can help organizations prepare such assessments and then provide help to design the facility.</p> <section class="section main-article-chapter" data-menu-title="What are the factors considered for data center tier classification?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What are the factors considered for data center tier classification?</h2> <p>The Uptime Institute assesses a data center facility or its design documents before allowing any organization to claim a tier classification. Tier Certification of Design Documents checks an organization's design documents against the Uptime Institute's criteria and then awards a tier classification based on that information.</p> <p>Tier Certification of Constructed Facility then ensures the <a href="https://www.datacenterknowledge.com/data-center-construction/2024-s-biggest-data-center-construction-stories-a-year-in-review">facility has been built accurately</a> against the design documents. Finally, Tier Certification of Operational Sustainability (TCOS) measures the capability of a facility to operate and meet expected availability and <a href="https://www.datacenterknowledge.com/build-design/unlocking-efficiency-the-advantages-of-multi-story-data-centers">efficiency</a> levels. All facilities with a data center tier classification are then listed on the Uptime Institute's website. There are currently more than 1,700 certifications across 98 countries.</p> <p>Tier classifications focus mainly on <a href="https://www.techtarget.com/searchitoperations/definition/platform">platform</a> availability. Areas of consideration include the following:</p> <ul class="default-list"> <li><b>Power.</b> What level of power redundancy is in place, not just within the data center, but from the grid to the facility?</li> <li><b>Cooling.</b> Does the <a href="https://www.techtarget.com/searchdatacenter/tip/Data-center-cooling-systems-and-technologies-and-how-they-work">facility have enough cooling</a> to meet the needs of critical areas of the platform? Is this cooling matched with <a href="https://www.techtarget.com/whatis/definition/redundancy">redundancy</a>? Is there any room for overhead?</li> <li><b>Maintenance.</b> Can the environment be maintained as planned and without significant, or any, downtime?</li> <li><b>Fault management capabilities.</b> Is the entire platform capable of surviving the failure of one or more components without affecting IT systems?</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="What are the different data center tiers?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What are the different data center tiers?</h2> <p>The Uptime Institute tiers are defined as follows.</p> <p><b>Tier I (basic capacity).</b> Tier I facilities protect against much of the human error involved in running <a href="https://www.techtarget.com/searchitoperations/feature/Data-center-operations-best-practices-checklist">data center operations</a>, but they aren't <a href="https://www.techtarget.com/searchdatacenter/definition/resiliency">resilient</a> to unexpected outages or equipment failure. Preventive maintenance and repairs generally require a complete shutdown of the IT capability. Lack of adequate planning to deal with maintenance on a regular basis could result in extended and costly platform downtime. Tier I includes the following components:</p> <ul class="default-list"> <li>Uninterruptible power supply (<a href="https://www.techtarget.com/searchdatacenter/definition/uninterruptible-power-supply">UPS</a>) to survive basic power outages, power sags and power spikes.</li> <li>Dedicated area for IT systems.</li> <li>Dedicated cooling systems that operate 24/7.</li> <li>Diesel or gas generators to cover longer power outages than the UPS can manage.</li> </ul> <p><b>Tier II (redundant capacity).</b> This classification builds on Tier I by adding more power resiliency through the following features:</p> <ul class="default-list"> <li>Engine generators, including fuel tanks to hold enough fuel to keep generators running for a longer period.</li> <li><a href="https://www.datacenterknowledge.com/uptime/how-bess-could-unlock-a-sustainable-future-for-data-centers">Energy storage</a> that enables the organization to switch from grid power services to engine power generation.</li> <li>Dedicated chillers or other cooling units, rather than simple air conditioning units.</li> <li>Heat rejection equipment that enables heat to be removed more effectively than with simple air conditioning.</li> </ul> <p><b>Tier III (concurrently maintainable).</b> The third tier builds on Tier II capabilities to provide enhanced platform availability, including these elements:</p> <ul class="default-list"> <li>Concurrent maintainability through the use of redundant components.</li> <li>Use of redundant power distribution paths.</li> </ul> <p><b>Tier IV (fault-tolerant).</b> These facilities provide the ultimate in platform resiliency by building on the previous tiers through the addition of multiple independent and isolated systems that provide <a href="https://www.techtarget.com/searchdatacenter/tip/Data-center-redundancy-The-basics">redundant capacity and power distribution</a>.</p> </section> <section class="section main-article-chapter" data-menu-title="How do the different tiers compare?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How do the different tiers compare?</h2> <p>Tier I data centers provide a competent environment with basic protection against systems failure and a degree of redundancy to manage power outages. However, these facilities don't have resilience when it comes to equipment failure.</p> <p>Tier II facilities provide much greater <a href="https://www.techtarget.com/searchdatacenter/tip/Data-center-backup-power-systems-standards-to-address-downtime">protection against power outages</a> through generation redundancies, as well as enhanced storage of fuel and possibly energy.</p> <p>Tier III facilities offer concurrent maintainability; this means the facility doesn't have to shut down to replace key equipment, as there is enough equipment redundancy to enable planned and unplanned maintenance.</p> <p>Tier IV facilities provide the ultimate in fault tolerance: Everything is dual-powered; redundant electrical, distribution and storage systems are available; and highly redundant server and <a href="https://www.techtarget.com/searchnetworking/tip/7-factors-to-consider-in-network-redundancy-design">network architectures</a> are in place. Tier IV facilities should survive any failure, provided there is enough fuel to fill the facility's generators and areas outside the data center owner's control, such as public networks, are available.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/whatis-uptime-data-center-tier-standards.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/whatis-uptime-data-center-tier-standards_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/whatis-uptime-data-center-tier-standards_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/whatis-uptime-data-center-tier-standards.png 1280w" alt="Table outlining the four uptime data center tier standards, their strategic approach and data center requirements." height="383" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>Data center tier standards outline what is required to ensure that reliability and performance needs are met. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> </section> <section class="section main-article-chapter" data-menu-title="Choosing the right data center tier"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Choosing the right data center tier</h2> <p>Many small and medium-sized businesses that have outgrown an office-based IT system -- a few servers and a simple PC/laptop network, for example -- find that a move to a Tier II, Tier III or Tier IV facility is expensive. A Tier I facility, which can be created by converting existing space within many office environments, is often enough in these situations.</p> <p>For those organizations where <a href="https://www.techtarget.com/searchnetworking/feature/The-true-cost-of-network-performance-issues-for-your-business">downtime is more of an issue</a> and where power outages are relatively common, a Tier II facility makes more sense, as it builds in more power fault tolerance.</p> <p>When the availability of the overall IT platform is more of an issue, Tier III offers systems redundancy with a strong capability to maintain a <a href="https://www.techtarget.com/searchdatacenter/definition/high-availability">high-availability</a> platform.</p> <figure class="main-article-image half-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/uptime_institute_uptime_guarantees-h.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/uptime_institute_uptime_guarantees-h_half_column_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/uptime_institute_uptime_guarantees-h_half_column_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/uptime_institute_uptime_guarantees-h.png 1280w" alt="Info box describing the Uptime Institute uptime guarantees across their four tiers." height="250" width="279"> <figcaption> <i class="icon pictures" data-icon="z"></i>Uptime guarantees are one way to differentiate data center tiers. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <p>Tier IV is for organizations where availability is non-negotiable -- for example, high-<a href="https://www.techtarget.com/searchnetworking/definition/throughput">throughput</a> financial systems or national security systems. While costly, this tier is a good investment, as equipment should always be available. However, Tier IV is based on single overall redundancy at all points, so the failure of one piece of equipment can become an opportunity for systems failure should a second item fail while the first is being replaced.</p> <p>To get around this issue, the Uptime Institute offers further certification levels using bronze, silver and gold TCOS levels. These levels recognize how facility owners have gone beyond the basic tier classifications -- for example, by adding three-way redundancy to their platforms.</p> <p>Data center owners who don't believe they can easily move the existing facility toward a tier-certified facility can take advantage of the Uptime Institute's professional services to see what would be required. The Institute also offers a separate service that performs an operational <a href="https://www.techtarget.com/searchcio/tip/What-is-risk-identification-Importance-and-methods">assessment</a> that can lead to a Management &amp; Operations Stamp of Approval rating.</p> </section> <section class="section main-article-chapter" data-menu-title="Alternatives to the Uptime Institute"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Alternatives to the Uptime Institute</h2> <p>The Uptime Institute's primary focal points in its tier standards are power systems, data center cooling, and <a href="https://www.techtarget.com/searchdatacenter/feature/Data-center-environmental-controls-a-high-priority-for-admins">environment management</a> and system availability. The standards have been popular for many years, owing largely to their simplicity and ease of understanding. That said, several alternatives have emerged in recent years that should be carefully examined, perhaps not to replace Uptime Institute standards but to identify additional criteria that can increase a data center's ability to perform.</p> <p>The following are current alternatives to the Uptime Institute's four tiers.</p> <p><b>ANSI/TIA-942 Telecommunications Infrastructure Standard for Data Centers.</b> Originally developed as a <a href="https://www.techtarget.com/searchnetworking/answer/What-are-the-6-components-of-structured-cabling">cabling</a> standard by the Telecommunications Industry Association (TIA), TIA-942 today addresses a broad range of data center attributes. It is recognized as an official American National Standards Institute (ANSI) standard, and its various elements offer important guidance when designing a new data center or preparing to upgrade one.</p> <p><b>ANSI/BICSI 002-2019 Data Center Design and Implementation Best Practices.</b> Often considered the most widely recognized data center design guideline, 002-2019 addresses the essential issues associated with data center design, from initial design and planning to commissioning. It includes design standards from other organizations, such as TIA and the <a href="https://www.techtarget.com/searchdatacenter/definition/ISO">International Organization for Standardization</a>.</p> <p><b>ISO 22237:2021 Information technology -- Data centre facilities and infrastructures.</b> Developed by the ISO, this standard provides a foundation for the design of data centers. It includes a classification system for data center criteria, plus descriptions of data center components and operational characteristics.</p> <p><b>Tier 5 Platinum standard from Switch Inc.</b> Developed to raise the bar over the Uptime Institute's four tiers, Tier 5 (not Roman numeral V) data centers exceed Tier IV criteria with the addition of components including enhanced power systems, multiple telecom carriers and power suppliers, positioning of cooling lines within the center, enhanced physical and <a href="https://www.techtarget.com/searchnetworking/answer/What-does-a-network-infrastructure-upgrade-project-involve">network infrastructure</a> security, and the use of fully renewable energy sources. The goal is a fully fault-tolerant data center and associated infrastructure. Switch also provides high-performance exascale data center ecosystems in several U.S. locations.</p> <p><b>The Green Grid data center standards.</b> Founded in 2007 to support the development of efficient data centers, The Green Grid was acquired by the Information Technology Industry Council in 2019. It offers the following tools to use when designing or updating data centers:</p> <ul class="default-list"> <li>TggCDU, a coolant distribution unit sizing tool.</li> <li>TggGPI, a performance indicator tool for assessing data center cooling.</li> <li>TggTCO, a total cost of operations calculator for cooling systems.</li> <li>TggWUI, a water usage impact calculation tool.</li> </ul> <p>The organization also provides the following resources for data center operators and developers:</p> <ul class="default-list"> <li><b>Data Center Resource Effectiveness (DCRE).</b> DCRE is a set of <a href="https://www.techtarget.com/searchdatacenter/feature/Words-to-go-6-key-data-center-sustainability-metrics">metrics</a> and a tool for examining multiple factors associated with data centers to ensure the center is energy efficient and environmentally friendly.</li> <li><b>Open Standard for Data Center Availability (OSDA).</b> Launched in 2016 to set a data center availability and rating system for energy-efficient and sustainable data center designs, OSDA is available as a document and online tool.</li> </ul> <div class="youtube-iframe-container"> <iframe id="ytplayer-0" src="https://www.youtube.com/embed/ucmnHYCawyA?autoplay=0&amp;modestbranding=1&amp;rel=0&amp;widget_referrer=null&amp;enablejsapi=1&amp;origin=https://www.techtarget.com" type="text/html" height="360" width="640" frameborder="0"></iframe> </div> </section> <section class="section main-article-chapter" data-menu-title="Pros and cons of Uptime Institute certification"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Pros and cons of Uptime Institute certification</h2> <p>Considering the ubiquity and acceptance of the Uptime Institute's tier standards, there are many reasons to obtain certification, including the Institute's reputation and credibility, getting certified to a recognized standard, ongoing identification of potential risks, support for continuous improvement and the prestige associated with an Uptime Institute certification.</p> <p>Conversely, the downsides of certification might include the costs and time needed to achieve certification, dealing with a complex certification process, and a limited scope of certification criteria -- e.g., uptime and reliability -- as compared to other standards that also address energy efficiency and sustainability.</p> <p><i>Data center tiers effectively categorize different types of facilities, but it is essential to recognize the ranking system's limitations. This understanding ensures informed decisions and optimal facility management. Learn more about the </i><a href="https://www.datacenterknowledge.com/uptime/weighing-the-pros-and-cons-of-data-center-tiers"><i>pros and cons of data center tiers</i></a><i>.</i></p> </section> Data center tiers are a system created by Uptime Institute LLC to describe the availability of infrastructure resources in a facility. https://cdn.ttgtmedia.com/visuals/digdeeper/1.jpg https://www.techtarget.com/searchdatacenter/definition/Uptime-data-center-tier-standards Mon, 16 Jun 2025 00:00:00 GMT <p>Power usage effectiveness (PUE) is a metric used to determine the energy efficiency of a <a href="https://www.techtarget.com/searchdatacenter/definition/data-center">data center</a>. PUE is determined by dividing the total amount of power entering a data center by the power used to run the IT equipment within it. PUE is expressed as a ratio, with overall efficiency improving as the quotient decreases toward 1.0.</p> <p>Data center <a href="https://www.techtarget.com/searchdatacenter/definition/infrastructure">infrastructure</a> and the processing power within it require a lot of energy, and data centers that do not operate efficiently will use more electricity. Monitoring a metric like PUE is useful for <a href="https://www.techtarget.com/searchcio/definition/benchmark">benchmarking</a> data center efficiency while keeping costs under control. Organizations and <a href="https://www.techtarget.com/searchdatacenter/tip/What-does-a-data-center-facility-manager-do">data center managers</a> can use this metric once to measure their data center efficiency using its energy consumption and then again to measure the effect of any changes made to the data center facility. This helps reduce power consumption and energy costs.</p> <p>PUE was created by members of <a target="_blank" href="https://www.thegreengrid.org/" rel="noopener">The Green Grid</a>, an industry group focused on data center energy efficiency. According to the "<a target="_blank" href="https://uptimeinstitute.com/resources/research-and-reports/uptime-institute-global-data-center-survey-results-2024" rel="noopener">Uptime Institute Global Data Center Survey Results 2024</a>" report, PUE and power consumption are among the top-tracked sustainability methods. A similar benchmarking standard recognized by The Green Grid is data center infrastructure efficiency, or <a href="https://www.techtarget.com/searchdatacenter/definition/data-center-infrastructure-efficiency-DCIE">DCiE</a>.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/power_usage_effectiveness-f.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/power_usage_effectiveness-f_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/power_usage_effectiveness-f_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/power_usage_effectiveness-f.png 1280w" alt="Diagram showing how power usage effectiveness (PUE) is calculated." height="280" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>The PUE calculation equation is straightforward, but it requires knowing the power requirements of deployed IT equipment and the power available for the data center. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <section class="section main-article-chapter" data-menu-title="How is PUE calculated?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How is PUE calculated?</h2> <p>The following equation calculates power usage effectiveness:</p> <p style="text-align: center;">PUE = Total facility power/IT equipment energy</p> <p><i>Total facility power</i> is the amount of operational power the facility uses, which includes all data center <a href="https://www.techtarget.com/searchnetworking/definition/hardware">hardware</a>, power delivery components, <a href="https://www.techtarget.com/searchdatacenter/tip/Data-center-cooling-systems-and-technologies-and-how-they-work">cooling systems</a> and lighting systems. <i>IT equipment energy</i> refers to the amount of energy that is used to power the <a href="https://www.techtarget.com/searchstorage/definition/storage">storage</a> and <a href="https://www.techtarget.com/searchnetworking/tip/An-introduction-to-8-types-of-network-devices">networking equipment</a> at the site as well as control equipment, such as monitors and workstations.</p> <p>As a value, PUE must be above 1.0 -- 1.3 to 3.0 is ideal -- which denotes efficient power delivery. Contributing factors can include minimal to no loss of electricity using the wiring infrastructure and equipment cooling achieved without the need for fans to move the air.</p> <p>The DCiE metric is often compared with power usage effectiveness. It is the inverse of PUE, as it divides IT equipment energy usage by the total energy used by the data center. For example, a PUE of 1.0 equals a DCiE value of 100% -- or 100% energy efficiency -- and a PUE of 2.0 equates to a DCiE of 50%.</p> <p>Calculate PUE using these steps:</p> <ol class="default-list"> <li><b>Measure energy use at the facility's utility meter.</b> If the data center is in a mixed-use facility or office building, take a measurement only at the meter that powers the data center. If it is not on a separate utility meter, estimate the amount of power being consumed by the non-data center portion of the building and remove it from the equation. A better way of doing this, however, is by using data center infrastructure management (<a href="https://www.techtarget.com/searchdatacenter/definition/data-center-infrastructure-management-DCIM">DCIM</a>) application software. Instead of checking the utility meter to make an estimation, additional sensors are installed that collect real-time energy usage data for measuring PUE.</li> <li><b>Measure the IT equipment load after power conversion, switching and conditioning are completed.</b> The most useful measurement point is at the <a href="https://www.techtarget.com/searchdatacenter/definition/power-distribution-unit-PDU">power distribution units</a>, which supply power to the computers and networking equipment. This measurement should represent the total power delivered to the <a href="https://www.techtarget.com/whatis/definition/rack-server-rack-mounted-server">server racks</a> in the data center.</li> <li><b>Do the math.</b> Once the measurements are collected, calculate PUE by dividing total facility power by equipment energy.</li> </ol> </section> <section class="section main-article-chapter" data-menu-title="Benefits and limitations of PUE"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Benefits and limitations of PUE</h2> <p>The benefits of measuring power usage effectiveness include the following:</p> <ul class="default-list"> <li><b>PUE calculation can be repeated.</b> The metric is meant to be used multiple times to gauge changes in data center efficiency.</li> <li><b>PUE value can gauge more efficient practices.</b> The metric can help identify effective ways to reduce energy usage.</li> <li><b>DCIM software integration.</b> DCIM software can automatically calculate PUE in <a href="https://www.techtarget.com/whatis/definition/real-time-monitoring">real time</a>, consolidating that data into reports.</li> <li><b>The PUE metric might be competitive.</b> Organizations with good PUE ratios can use their score in marketing materials.</li> </ul> <p>However, there are some limitations involved with data center PUE:</p> <ul class="default-list"> <li><b>Total facility power calculation is not always accurate.</b> The accuracy of the total facility power calculation is a major factor affecting the measurement of PUE. If an organization has to estimate that number from a shared utility meter, then the resulting calculation is only an estimation.</li> <li><b>PUE does not capture consumption at the rack level. </b> Energy that is consumed at the rack level is not part of the overall PUE calculation, which might make the PUE less accurate than desired.</li> <li><b>PUE is not meant as a marketing metric.</b> Although organizations can use PUE in marketing, the metric is meant to be used as an improvement and assessment metric. Marketers should not rely heavily on PUE just because it might be a widely understood concept.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="How can you lower PUE?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How can you lower PUE?</h2> <p>The following steps can help lower a data center's power usage effectiveness ratio closer to 1.0:</p> <ul class="default-list"> <li><b>Virtualize servers.</b> Multiple <a href="https://www.techtarget.com/searchitoperations/definition/host-virtual-machine-host-VM">virtual machines</a> can run their own workloads on a single hardware server, which reduces energy consumption and frees up floor space.</li> <li><b>Improve cooling systems.</b> Data centers <a href="https://www.techtarget.com/searchdatacenter/tip/Evaluate-air-cooling-vs-liquid-cooling-for-your-data-center">require cooling systems</a> to prevent overheating. However, refrigerant-based cooling systems use a lot of power. Improving these systems or reducing the data center's reliance on them can help lower PUE.</li> <li><b>Optimize cool air production.</b> This can be done, for example, by using naturally cool outside air or heat exchangers, rather than relying on refrigerants.</li> <li><b>Replace inefficient hardware.</b> The quality and performance of some hardware might degrade over time, so if servers or storage systems are not performing properly, they should be replaced.</li> <li><b>Use an energy-efficient uninterruptible power supply (UPS).</b> Power distribution should be designed with a <a href="https://www.techtarget.com/searchdatacenter/definition/uninterruptible-power-supply">UPS</a> to be more efficient. More efficient equipment and making power run a shorter distance help increase efficiency.</li> <li><b>Use energy-efficient lighting.</b> Although lighting generally makes up a small portion of energy consumption, it is still an easy way to reduce power and heat production. Replacing fluorescent lighting with LEDs on motion <a href="https://www.techtarget.com/whatis/definition/sensor">sensors</a> and lighting controls can help reduce operational power consumption and heat production.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="Power usage effectiveness system manufacturers"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Power usage effectiveness system manufacturers</h2> <p>PUE measurement technology is often provided by vendors who manufacture systems that examine and analyze overall data center infrastructure efficiency. The following are among the types of firms offering PUE measurement and expertise:</p> <ul class="default-list"> <li><strong>Cooling system providers.</strong> Such firms provide systems and expertise for advanced cooling technologies that can reduce energy usage.</li> <li><strong>Data center infrastructure management tools.</strong> DCIM systems handle the analysis and management of the many issues associated with a data center infrastructure.</li> <li><strong>IT equipment manufacturers.</strong> Companies such as Dell and HPE offer energy-efficient servers and storage devices.</li> <li><strong>Consulting Services</strong>. The Uptime Institute and The Green Grid provide consultancy services to improve PUE.</li> </ul> <p>These vendors are among the companies that offer DCIM products:</p> <ul class="default-list"> <li>Device42.</li> <li>Hyperview.</li> <li>IBM Maximo.</li> <li>ManageEngine.</li> <li>Microsoft.</li> <li>Modius.</li> <li>Nlyte Software.</li> <li>Schneider Electric.</li> <li>Sunbird Software.</li> <li>Vertiv.</li> </ul> <p>When selecting firms whose systems address PUE measurement, check each vendor's capabilities carefully to ensure they meet your infrastructure requirements.</p> <div class="youtube-iframe-container"> <iframe id="ytplayer-0" src="https://www.youtube.com/embed/yGkfBo2iSiI?autoplay=0&amp;modestbranding=1&amp;rel=0&amp;widget_referrer=null&amp;enablejsapi=1&amp;origin=https://www.techtarget.com" type="text/html" height="360" width="640" frameborder="0"></iframe> </div> </section> <section class="section main-article-chapter" data-menu-title="Impact of AI on power usage effectiveness"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Impact of AI on power usage effectiveness</h2> <p>As with most technologies in use today, <a href="https://www.techtarget.com/searchenterpriseai/definition/AI-Artificial-Intelligence">AI</a> greatly enhances the capabilities of data center infrastructure and energy management technologies. AI can enhance and optimize energy use by automating various repetitive tasks, such as power usage, cooling system performance monitoring and maintenance scheduling. Ironically, the use of AI applications can also increase costs related to greater energy demand, which can negatively affect the data center's carbon footprint. Ultimately, the use of AI for power management can help balance energy usage and environmental concerns.</p> <p><i>The increasing carbon footprint of artificial intelligence is alarming. Discover </i><a href="https://www.techtarget.com/sustainability/feature/How-CIOs-can-reduce-AIs-negative-environmental-impact"><i>how leaders can promote energy efficiency and sustainability for AI</i></a><i>. Also, UPS systems typically require continuous maintenance over time. Learn more </i><a href="https://www.techtarget.com/searchdatacenter/tip/Safeguard-data-center-power-with-regular-UPS-maintenance"><i>about UPS maintenance</i></a><i> and how it helps safeguard data center power.</i></p> </section> Power usage effectiveness (PUE) is a metric used to determine the energy efficiency of a data center. https://cdn.ttgtmedia.com/visuals/digdeeper/5.jpg https://www.techtarget.com/searchdatacenter/definition/power-usage-effectiveness-PUE Thu, 12 Jun 2025 00:00:00 GMT <p>A computer room air conditioning (CRAC) unit is a device that monitors and maintains the temperature, air distribution and humidity in a <a href="https://www.techtarget.com/searchdatacenter/definition/data-center">data center</a>, network or server room. CRAC units replace the air conditioning (AC) units used in the past to cool data centers. By comparison, CRAC systems have better air filtration, better humidity control mechanisms and higher airflow than typical AC systems.</p> <p>CRAC units, which are a type of heating, ventilation and air conditioning (<a href="https://www.techtarget.com/searchdatacenter/definition/HVAC">HVAC</a>) system, help prevent low humidity levels and water vapor from forming. Low humidity can cause static electricity buildup, which can result in electrostatic discharge (<a href="https://www.techtarget.com/whatis/definition/electrostatic-discharge-ESD">ESD</a>) that could damage electronics in computer equipment, and water vapor buildup, which can cause short circuits and corrode equipment. CRAC units are generally more efficient than air conditioning systems and are designed for medium to large data centers. It is a good idea, though, to have portable air conditioning units available in case of a CRAC system failure or scheduled downtime.</p> <p>Computer room air conditioning units work using a <a href="https://www.techtarget.com/searchdatacenter/definition/Mechanical-refrigeration">mechanical refrigeration</a> cycle where air is blown over a cooling coil filled with refrigerant. Refrigerant in the cooling coil is kept cold by a compressor.</p> <p>Excess heat is expelled as air, water or a glycol mixture. Older computer room air conditioners can only turn on and off, while newer units enable different airflow variations and have greater energy efficiency and cooling capacity.</p> <p>There are various ways that CRAC units can be situated. One popular CRAC setup is having cooling air dispensed through an elevated floor. Also called <a href="https://www.techtarget.com/searchdatacenter/definition/raised-floor">raised flooring</a>, the cooled air rises through the perforated floor sections, forming cold aisles. The cold air flows through the racks, where it picks up heat before exiting from the rear of the racks. The warm exit air forms <a href="https://www.techtarget.com/searchdatacenter/tip/Explore-hot-and-cold-aisle-containment-for-your-data-center">hot aisles</a> behind the racks, and the hot air returns to the CRAC intakes, which are positioned above the floor.</p> <figure class="main-article-image half-col" data-img-fullsize="https://www.techtarget.com/rms/dataCenter-Virtualization/CRAC_Hot_Cold_Aisle.png"> <img data-src="https://www.techtarget.com/rms/dataCenter-Virtualization/CRAC_Hot_Cold_Aisle_half_column_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/dataCenter-Virtualization/CRAC_Hot_Cold_Aisle_half_column_mobile.png 960w,https://www.techtarget.com/rms/dataCenter-Virtualization/CRAC_Hot_Cold_Aisle.png 1280w" alt="Hot Aisle/Cold Aisle approach diagram."> <figcaption> <i class="icon pictures" data-icon="z"></i>Computer air conditioning systems often employ a hot aisle/cold aisle approach to keep cool air moving from beneath the raised floor, through the racks and into the system's intake vents. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <section class="section main-article-chapter" data-menu-title="Components of a CRAC unit"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Components of a CRAC unit</h2> <p>CRAC units include the following components:</p> <ul class="default-list"> <li><b>Refrigerant.</b> This is a chemical compound that is used as a coolant in CRAC and air conditioners.</li> <li><b>Cooling coils.</b> Refrigerant flows through the cooling coils -- also called <i>evaporator coils</i> -- to cool the air.</li> <li><b>Compressors.</b> The compressor keeps the refrigerant running through the coil cool.</li> <li><b>Air filters.</b> These filters help catch the accumulation of dirt or dust that travels through a CRAC unit.</li> <li><b>Electronically commutated fans.</b> EC fans blow cool air while producing less heat with high efficiency.</li> <li><b>Humidifiers and dehumidifiers.</b> Although these are optional in CRAC systems, if needed, humidifiers increase the amount of moisture in the air, while dehumidifiers remove water vapor from the air.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="What are the differences between CRAC and CRAH units?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What are the differences between CRAC and CRAH units?</h2> <p>Computer room air handler (<a href="https://www.techtarget.com/searchdatacenter/definition/computer-room-air-handler-CRAH">CRAH</a>) units are another data center cooling device. The main difference between CRAH and CRAC units is how they cool their surrounding environments. CRAC units use both refrigerants and compressors, while CRAH units use chilled water and control valves.</p> <p>CRAH units use fans to blow air over cooling coils. Instead of the coils being filled with refrigerant, however, they're filled with chilled water. Warm air from the data center room is drawn back into the CRAH unit, passing over the chilled water coils. The heat transfers from the air to the water, where the water then returns to a chiller.</p> <p>Because CRAH units don't need a compressor, they consume less energy and are more efficient than CRAC units. This makes them more suitable for <a href="https://www.techtarget.com/searchdatacenter/tip/Considerations-for-sustainable-data-center-design">sustainable data centers</a>.</p> <p>CRAC units are ideal for data centers with electrical loads of 200 kilowatts (kW) or less. Meanwhile, CRAH units are designed for data centers with electrical loads of 200 kW or more. CRAH units are also ideal for environments with <a href="https://www.techtarget.com/searchdatacenter/definition/high-availability">high availability</a> requirements.</p> </section> <section class="section main-article-chapter" data-menu-title="Pros and cons of computer room air conditioning"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Pros and cons of computer room air conditioning</h2> <p>Because the equipment generally found in data centers generates a lot of heat, some sort of cooling is essential to maintain the proper temperature and humidity. As noted earlier, CRAC units establish a proper environment for equipment by providing the recommended air filtration, temperature and humidity control, and airflow. This minimizes static electricity and water vapor buildup, both of which can damage equipment.</p> <p>Challenges include costs for a CRAC system, such as the following:</p> <ul class="default-list"> <li>Design, installation, testing and maintenance.</li> <li>Managing system operation to ensure the correct operating environment can be maintained.</li> <li>Having resources to regularly test the ambient environment to ensure optimum CRAC performance.</li> </ul> <div class="youtube-iframe-container"> <iframe id="ytplayer-0" src="https://www.youtube.com/embed/8fjdpJ-9JI4?autoplay=0&amp;modestbranding=1&amp;rel=0&amp;widget_referrer=null&amp;enablejsapi=1&amp;origin=https://www.techtarget.com" type="text/html" height="360" width="640" frameborder="0"></iframe> </div> </section> <section class="section main-article-chapter" data-menu-title="Advanced technology for data center cooling"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Advanced technology for data center cooling</h2> <p>Today's data centers take advantage of the many advancements in data center environment and operations management. Among these are smart assistants that use <a href="https://www.techtarget.com/searchenterpriseai/definition/AI-Artificial-Intelligence">AI</a> and machine learning (<a href="https://www.techtarget.com/searchenterpriseai/definition/machine-learning-ML">ML</a>) to <a href="https://www.techtarget.com/searchdatacenter/answer/How-can-I-build-AI-capabilities-for-the-data-center">monitor facilities</a> efficiently and activate adjustments to HVAC systems to maintain temperature and humidity levels. AI-based HVAC management of heat output and cooling usage can result in energy savings.</p> <p><a href="https://www.techtarget.com/whatis/definition/real-time-monitoring">Real-time monitoring</a> of temperature and humidity level can be aided by robots that systematically move throughout the data center, capturing equipment cabinet and <a href="https://www.techtarget.com/whatis/definition/rack">rack</a> temperatures and humidity levels.</p> </section> <section class="section main-article-chapter" data-menu-title="CRAC system planning considerations"> <h2 class="section-title"><i class="icon" data-icon="1"></i>CRAC system planning considerations</h2> <p>Building or updating a data center is a complex task. When it comes to heating and cooling, the following steps should be part of the planning process:</p> <ul class="default-list"> <li>Examine existing systems to see if their energy efficiency is appropriate for the currently installed and new equipment.</li> <li>New data centers will need to start at the design stage when determining heating and cooling requirements.</li> <li>Identify situations where existing components, such as thermostats, are not performing properly.</li> <li>Identify and work with experienced HVAC engineers when developing the design for a new data center or updating an existing one.</li> <li>If a chilled water facility is needed, the HVAC engineer can provide guidance on a suitable location.</li> <li>Vendors with experience in <a href="https://www.techtarget.com/searchdatacenter/How-to-design-and-build-a-data-center">data center design</a> should be considered, especially when configuring underfloor air flow and piping for water and/or coolant delivery.</li> <li>The HVAC plant should be scalable to accommodate changes in data center operations.</li> <li>Ensure that the primary CRAC system has suitable backups, including emergency power, to ensure uninterrupted operation.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="Manufacturers of computer room air conditioning systems"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Manufacturers of computer room air conditioning systems</h2> <p>The following is a list of CRAC system manufacturers:</p> <ul class="default-list"> <li>Airedale by Modine.</li> <li>Green Revolution Cooling.</li> <li>Hitachi.</li> <li>Mee Industries.</li> <li>Mitsubishi Electric.</li> <li>Nortek Air Solutions.</li> <li>Rittal.</li> <li>Schneider Electric.</li> <li>Silent-Aire.</li> <li>Stulz.</li> <li>Vertiv.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="The future of data center cooling system technologies"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The future of data center cooling system technologies</h2> <p>As data centers increase in complexity and size, HVAC technologies will need to keep in step to optimize data center performance. Liquid cooling systems using water or coolants are widely used today, and new data center cooling technologies are likely to appear in the coming years. These can include <a href="https://www.techtarget.com/whatis/definition/geothermal-cooling">geothermal cooling</a> and smart technologies using AI to more efficiently and environmentally monitor and manage cooling activities.</p> <p><i>Learn more about the cooling systems used in data centers and</i> <i><a href="https://www.techtarget.com/searchdatacenter/feature/Liquid-cooling-vs-air-cooling-in-the-data-center">how liquid and air cooling compare</a>.</i></p> </section> A computer room air conditioning (CRAC) unit is a device that monitors and maintains the temperature, air distribution and humidity in a data center, network or server room. https://cdn.ttgtmedia.com/visuals/digdeeper/5.jpg https://www.techtarget.com/searchdatacenter/definition/computer-room-air-conditioning-unit Tue, 10 Jun 2025 09:00:00 GMT <p>Carbon footprint measures the total <a href="https://www.techtarget.com/whatis/definition/greenhouse-gas">greenhouse gas</a> emissions caused directly and indirectly by an individual, organization, product or activity. It's typically expressed in metric tons of carbon dioxide equivalent per year; CO2e is a standardized unit that combines the global warming potential of multiple greenhouse gases -- including carbon dioxide, methane and nitrous oxide -- into a single comparable metric.</p> <p>Originally intended to measure all emissions contributing to global warming, the term <i>carbon footprint</i> is still commonly used as shorthand for all greenhouse gas emissions. <a href="https://www.informationweek.com/sustainability/carbon-labeling-and-emission-tracking-takes-hold-with-consumers">Tracking a carbon footprint</a> is one of the most accessible ways to understand and mitigate climate change impacts.</p> <section class="section main-article-chapter" data-menu-title="Why carbon footprints matter"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Why carbon footprints matter</h2> <p>Understanding and <a href="https://www.techtarget.com/sustainability/feature/Sustainability-management-software-providers-to-consider">managing carbon footprints</a>&nbsp;is critical to addressing human-caused climate change. The more greenhouse gases emitted, the more they trap heat in Earth's atmosphere, resulting in rising temperatures, more frequent extreme weather events and ecosystem disruption.</p> <p>Reducing carbon footprints at the individual and organizational level is key to meeting global emission reduction targets outlined in agreements like the Paris Agreement and COP28 commitments. However, before individuals or businesses can hope to <a href="https://www.techtarget.com/sustainability/feature/ESG-strategy-and-management-Complete-guide-for-businesses">reduce their carbon footprint</a>, it's key to understand how emissions are categorized.</p> <h3>Primary vs. secondary carbon footprints</h3> <p>A carbon footprint is composed of two parts, a primary and secondary footprint.</p> <p>The primary footprint includes direct emissions from activities such as the following:</p> <ul class="default-list"> <li>Driving gas-powered vehicles.</li> <li>Heating a home with oil, natural gas or coal.</li> <li>Air travel.</li> <li>Electricity usage when derived from fossil fuels.</li> </ul> <p>These emissions result directly from the combustion of fossil fuels.</p> <p>The secondary footprint refers to indirect emissions from the following:</p> <ul class="default-list"> <li>Manufacturing and transporting products and food.</li> <li>Industrial <a href="https://www.techtarget.com/whatis/definition/supply-chain">supply chains</a>.</li> <li>Use of digital services and <a href="https://www.techtarget.com/searchcloudcomputing/definition/cloud-infrastructure">cloud infrastructure</a>.</li> <li>Waste and recycling processes.</li> </ul> <p>This includes the full lifecycle emissions of goods and services consumed, from production to disposal.</p> </section> <section class="section main-article-chapter" data-menu-title="How to reduce carbon footprints"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How to reduce carbon footprints</h2> <p>Reducing carbon footprints has become a major focus in both personal and <a href="https://www.techtarget.com/sustainability/feature/CIOs-turn-to-ESG-tech-as-part-of-sustainability-leadership">corporate sustainability initiatives</a>. Key strategies include the following.</p> <h3>Switch to renewable energy</h3> <p>Powering homes or businesses with solar, wind or hydroelectric power can eliminate fossil fuel-based emissions from electricity use.</p> <h3>Choose low-emissions transportation</h3> <p>Use public transit, cycle, walk or switch to electric vehicles to reduce transport-related emissions.</p> <h3>Adopt sustainable consumption habits</h3> <p>Take note of personal and professional behaviors and, where possible, swap them for more <a href="https://www.techtarget.com/whatis/feature/ESG-vs-CSR-vs-sustainability-Whats-the-difference">sustainable alternatives</a>. Below are a few examples:</p> <ul class="default-list"> <li>Buy local and seasonal food.</li> <li>Reduce meat and dairy consumption.</li> <li>Avoid single-use plastics.</li> <li>Support circular economy practices.</li> </ul> <h3>Implement green building practices</h3> <p>Energy-efficient construction, sustainable insulation, smart thermostats and low-carbon building materials help <a href="https://omdia.tech.informa.com/blogs/2024/apr/the-ict-industry-is-a-force-for-good">reduce the environmental effects</a> of buildings.</p> <h3>Purchase carbon offsets</h3> <p><a href="https://www.techtarget.com/whatis/definition/carbon-offset">Carbon offsets</a> are credits used to compensate for emissions by funding environmental projects like reforestation, wind farms or methane capture. While not a long-term solution, they help organizations reduce net emissions when direct reductions are not possible.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/how_to_offset_carbon_emissions-f.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/how_to_offset_carbon_emissions-f_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/how_to_offset_carbon_emissions-f_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/how_to_offset_carbon_emissions-f.png 1280w" alt="Graphic showing how to offset carbon emissions." height="336" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>To reduce greenhouse gas emissions and meet emissions reduction targets, one strategy is to offset carbon emissions -- follow these four steps. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <p>Several countries worldwide have set targets for reductions in emissions, referred to as carbon reduction commitments, in international meetings with agreements like the Paris Agreement, the Copenhagen Accord and the Kyoto Protocol.</p> </section> <section class="section main-article-chapter" data-menu-title="Global climate targets and carbon commitments"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Global climate targets and carbon commitments</h2> <p>Numerous international frameworks aim to limit global warming to 1.5 degrees Celsius above preindustrial levels, a goal set by the Paris Agreement. Key climate accords that define carbon reduction efforts include the following:</p> <ul class="default-list"> <li><b>Paris Agreement (2015).</b> Signed by 196 countries to set legally binding climate goals.</li> <li><b>COP28 (2023).</b> Called for phasing out unabated fossil fuels and tripling renewable energy capacity by 2030.</li> <li><b>Kyoto Protocol (1997).</b> One of the first international efforts to establish emissions reduction targets.</li> <li><b>Net-zero pledges.</b> Many countries and companies have committed to achieving <a href="https://www.techtarget.com/whatis/definition/net-zero">net-zero</a> carbon emissions by 2050.</li> </ul> <p>Governments are also creating carbon pricing schemes, such as carbon taxes and <a target="_blank" href="https://www.epa.gov/emissions-trading/what-emissions-trading" rel="noopener">emissions trading</a> systems, to incentivize businesses to cut emissions.</p> </section> <section class="section main-article-chapter" data-menu-title="New trends in carbon measurement"> <h2 class="section-title"><i class="icon" data-icon="1"></i>New trends in carbon measurement</h2> <p>As our climate continues to change and evolve, so must our response to protect it. With that mindset, modern carbon tracking has been designed to include the following:</p> <ul class="default-list"> <li><a href="https://www.techtarget.com/searchenterpriseai/definition/AI-Artificial-Intelligence">AI</a>-powered carbon calculators, or personalized apps that calculate and suggest reduction actions.</li> <li>Product-level carbon labels on packaging, which show carbon footprints to inform consumers.</li> <li><a href="https://www.informationweek.com/sustainability/what-are-scope-3-emissions-and-how-can-data-centers-address-them-">Scope 3 emissions</a> tracking, as companies are increasingly required to report indirect emissions from suppliers, logistics and product use.</li> <li>Digital carbon twins, or real-time models that simulate an organization's carbon output across its entire operation.</li> </ul> <p>These advances make it easier than ever for individuals and businesses to measure and reduce their environmental impacts in meaningful ways.</p> <div class="youtube-iframe-container"> <iframe id="ytplayer-0" src="https://www.youtube.com/embed/8de2fUKq77o?autoplay=0&amp;modestbranding=1&amp;rel=0&amp;widget_referrer=null&amp;enablejsapi=1&amp;origin=https://www.techtarget.com" type="text/html" height="360" width="640" frameborder="0"></iframe> </div> <p><em>Businesses are essential in the fight against climate change. Explore <a href="https://www.techtarget.com/whatis/feature/Ways-businesses-can-reduce-their-carbon-footprint">ways businesses can reduce their carbon footprint</a>. Also, learn about the <a href="https://www.techtarget.com/sustainability/feature/A-simple-guide-to-greenhouse-gases-Main-types-and-sources">main types of greenhouse gases</a>, how <a href="https://www.techtarget.com/virtualhealthcare/news/366623086/Evidence-that-telehealth-cuts-carbon-emissions-grows">evidence shows that telehealth cuts carbon emissions</a> and about <a href="https://www.techtarget.com/sustainability/feature/Generative-AIs-sustainability-problems-explained">generative AI's sustainability problems</a>.</em></p> </section> A carbon footprint measures the total greenhouse gas emissions caused directly and indirectly by an individual, organization, product or activity. https://cdn.ttgtmedia.com/visuals/digdeeper/3.jpg https://www.techtarget.com/whatis/definition/carbon-footprint Tue, 03 Jun 2025 12:00:00 GMT <p>Total cost of ownership (TCO) is an estimation of the expenses associated with purchasing, deploying, managing, using and retiring <a href="https://www.techtarget.com/searchcio/definition/IT-asset-management-information-technology-asset-management">IT assets</a>, such as a product or piece of equipment.</p> <p>TCO, or actual cost, quantifies the cost of the purchase across the product's entire lifecycle. Therefore, it offers a more accurate basis for determining the value -- cost vs. return on investment (<a href="https://www.techtarget.com/searchcio/definition/ROI">ROI</a>) -- of an investment than the purchase price alone.</p> <p>TCO can be calculated as the initial purchase price plus costs of operation across the asset's lifespan. It is especially critical in <a href="https://www.techtarget.com/searchdatacenter/definition/IT">IT</a>, manufacturing, <a href="https://www.techtarget.com/searcherp/definition/supply-chain-management-SCM">supply chain management</a> and <a href="https://www.techtarget.com/searchcloudcomputing/definition/cloud-computing">cloud computing</a>, where operational costs often exceed initial purchase costs.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineImages/ROI_example.jpg"> <img data-src="https://www.techtarget.com/rms/onlineImages/ROI_example_mobile.jpg" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineImages/ROI_example_mobile.jpg 960w,https://www.techtarget.com/rms/onlineImages/ROI_example.jpg 1280w" alt="graphic illustrating how to calculate return on investment in an acquisition" height="226" width="519"> <figcaption> <i class="icon pictures" data-icon="z"></i>Total cost of ownership -- aka money spent -- is a factor to consider when determining return on investment. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <section class="section main-article-chapter" data-menu-title="What factors determine TCO?"> <h2 class="section-title"><i class="icon" data-icon="1"></i><a name="_gb22my6sfg0k"></a>What factors determine TCO?</h2> <p>Overall TCO includes direct and indirect expenses, as well as some intangible ones that may be assigned a monetary value. TCO includes the following:</p> <ul class="default-list"> <li>Direct costs, e.g., purchase price, installation, maintenance.</li> <li>Indirect costs, e.g., training, downtime, performance inefficiencies.</li> <li>Intangible costs, e.g., employee productivity loss, compliance risks.</li> </ul> <p>For example, a <a href="https://www.techtarget.com/whatis/definition/server">server</a>'s TCO might include an expensive purchase price. However, indirect costs could include a good deal on ongoing IT support and also factor in low <a href="https://www.techtarget.com/searchitoperations/definition/systems-management">systems management</a> time because of its user-friendly interface.</p> <p>TCO factors in the costs accumulated from purchase to decommissioning of the asset.</p> <p>For a <a href="https://www.techtarget.com/searchdatacenter/definition/data-center">data center</a> server, for example, this means initial acquisition price, repairs, maintenance costs, upgrades, service or support contracts, network integration, security, <a href="https://www.techtarget.com/searchcio/definition/software-license">software licenses</a> and employee training.</p> <p>It can even account for the credit terms on which the company purchased the product. Through analysis, the purchasing manager might assign a monetary value to intangible costs, such as systems management time, electricity used, <a href="https://www.techtarget.com/whatis/definition/uptime-and-downtime">downtime</a>, insurance and other overhead.</p> <p>Total cost of ownership must be compared to total benefits of ownership to determine the viability of a purchase.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/cio-it_asset_lifecycle_management.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/cio-it_asset_lifecycle_management_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/cio-it_asset_lifecycle_management_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/cio-it_asset_lifecycle_management.png 1280w" alt="graphic illustrating the lifecycle of an IT asset from planning to retiring the asset" height="560" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>Total cost of ownership of an IT product covers not just the initial purchase price, but costs of operating that product across its lifespan -- when it is retired by an organization. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <h3><a name="_wkzp0xj81m3s"></a>New TCO considerations today</h3> <p>In addition to the aforementioned criteria, total cost of ownership today also may include the following:</p> <ul class="default-list"> <li>Cloud computing and software-as-a service (<a href="https://www.techtarget.com/searchcloudcomputing/definition/Software-as-a-Service">SaaS</a>) costs <ul style="list-style-type: circle;" class="default-list"> <li>SaaS subscription models introduce unpredictable long-term costs.</li> <li>Hidden fees for data storage, <a href="https://www.techtarget.com/searchapparchitecture/definition/application-program-interface-API">application programming interface</a> calls and <a href="https://www.techtarget.com/searchdatacenter/definition/vendor-lock-in">vendor lock-in</a> affect TCO.</li> <li><a href="https://www.techtarget.com/searchcloudcomputing/tip/8-key-steps-of-a-cloud-exit-strategy">Cloud egress</a> fees -- costs to move data out of a cloud provider -- are often overlooked.</li> </ul> </li> <li>Artificial intelligence (<a href="https://www.techtarget.com/searchenterpriseai/definition/AI-Artificial-Intelligence">AI</a>)-powered cost optimization <ul style="list-style-type: circle;" class="default-list"> <li>AI can help to predict maintenance costs and optimize <a href="https://www.techtarget.com/searchdatacenter/definition/infrastructure">IT infrastructure</a> spending.</li> <li><a href="https://www.techtarget.com/searchbusinessanalytics/definition/predictive-analytics">Predictive analytics</a> helps forecast vendor pricing changes.</li> </ul> </li> <li><a href="https://www.techtarget.com/whatis/definition/business-sustainability">Sustainability</a> and <a href="https://www.techtarget.com/searchcio/definition/green-IT-green-information-technology">green IT</a> <ul style="list-style-type: circle;" class="default-list"> <li>Energy-efficient servers, cloud data centers and carbon footprint reductions affect TCO.</li> <li>Regulatory requirements for sustainability reporting add compliance costs.</li> </ul> </li> </ul> </section> <section class="section main-article-chapter" data-menu-title="The challenges in calculating TCO"> <h2 class="section-title"><i class="icon" data-icon="1"></i><a name="_evpq73ihsncd"></a>The challenges in calculating TCO</h2> <p>There are several methodologies and <a href="https://www.techtarget.com/searchapparchitecture/definition/software">software</a> tools to calculate total cost of ownership, but the process is not perfect. Many enterprises fail to define a singular methodology. This is an issue because they cannot base purchasing decisions on uniform information.</p> <p>Another problem is that it is difficult to determine the scope of <a href="https://www.techtarget.com/whatis/definition/OPEX-operational-expenditure">operating costs</a> for any single piece of IT equipment. Some hidden cost factors are easily overlooked, such as depreciation and warranty, or inaccurately compared from one product to another.</p> <p>For example, support costs on one server may include the cost of spare parts. This might make support cost more than it does on another server but eliminate the acquisition cost of buying a new system.</p> <p>Cost of ownership analysis generally doesn't anticipate unpredictable rising costs over time -- for example, if <a href="https://www.techtarget.com/searchnetworking/tip/How-to-plan-and-start-a-network-upgrade">upgrade</a> part costs jump substantially more than expected due to a distributor change.</p> <p>TCO calculations cannot account for the availability of upgrades and services or the impact of vendor relationships.</p> <p>If a software vendor cancels a particular functionality after three years, no longer stocks parts after five years or ends support for certain software, the enterprise may be subject to unexpected and significant additional costs, which could drive TCO far beyond its initial estimate.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineImages/converged_infras-hci_tco_checklist-f.png"> <img data-src="https://www.techtarget.com/rms/onlineImages/converged_infras-hci_tco_checklist-f_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineImages/converged_infras-hci_tco_checklist-f_mobile.png 960w,https://www.techtarget.com/rms/onlineImages/converged_infras-hci_tco_checklist-f.png 1280w" alt="Text graphic outlining the various cost elements associated with an infrastructure project" height="498" width="559"> <figcaption> <i class="icon pictures" data-icon="z"></i>To understand the range of factors that go into any TCO calculation, consider what goes into TCO for hyperconverged infrastructure when compute, network and storage are bundled in a single platform. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> </section> <section class="section main-article-chapter" data-menu-title="TCO in cloud vs. on-premises IT infrastructure"> <h2 class="section-title"><i class="icon" data-icon="1"></i><a name="_6ahm1p82matk"></a>TCO in cloud vs. on-premises IT infrastructure</h2> <p>Given the information covered in the previous section, the following serves as a breakdown of TCO factors in a <a href="https://www.techtarget.com/searchcloudcomputing/tip/Evaluate-on-premises-vs-cloud-computing-pros-and-cons">cloud versus on-premises</a> infrastructure.</p> <table class="main-article-table"> <thead> <tr> <td>Factor</td> <td>Cloud (SaaS, IaaS, PaaS)</td> <td>On-premises (servers, data centers)</td> </tr> </thead> <tbody> <tr> <td><strong>Upfront cost</strong></td> <td>Low (subscription-based)</td> <td>High (Capex investment)</td> </tr> <tr> <td><strong>Maintenance cost</strong></td> <td>Managed by provider</td> <td>In-house IT team</td> </tr> <tr> <td><strong>Scalability</strong></td> <td>High (elastic)</td> <td>Limited</td> </tr> <tr> <td><strong>Energy cost</strong></td> <td>Included in fees</td> <td>Paid separately</td> </tr> <tr> <td><strong>Vendor lock-in risk</strong></td> <td>High</td> <td>Low</td> </tr> <tr> <td><strong>Long-term TCO</strong></td> <td>Variable</td> <td>Predictable</td> </tr> </tbody> </table> </section> <section class="section main-article-chapter" data-menu-title="Best practices to optimize TCO calculations"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Best practices to optimize TCO calculations</h2> <p>Enterprise managers and purchasing decision-makers complete cost analyses for multiple options and then compare TCO to determine overall costs and, ultimately, the lowest long-term cost. The following <a href="https://www.techtarget.com/searchsoftwarequality/definition/best-practice">best practices</a> can help organizations optimize TCO:</p> <ul class="default-list"> <li><a name="_xvdr86npgg5z"></a>Use AI and predictive analytics <ul style="list-style-type: circle;" class="default-list"> <li>AI-powered tools can analyze historical spending patterns and forecast future costs.</li> <li>AI-driven <a href="https://www.techtarget.com/searchcloudcomputing/tip/Cloud-cost-management-tools-you-should-know-about">cloud cost management platforms</a> -- e.g., Apptio, Cloudability -- help optimize cloud TCO.</li> </ul> </li> <li><a name="_tl521w723gwf"></a>Factor in <a href="https://www.techtarget.com/whatis/definition/environmental-social-and-governance-ESG">environmental, sustainability and governance</a> costs <ul style="list-style-type: circle;" class="default-list"> <li>Track <a href="https://www.techtarget.com/sustainability/feature/Understand-greenhouse-gas-emissions-vs-carbon-emissions">carbon emissions</a> from IT infrastructure.</li> <li>Invest in <a href="https://www.techtarget.com/searchdatacenter/tip/Four-ways-to-reduce-data-center-power-consumption">energy-efficient</a> hardware to lower long-term operational costs.</li> </ul> </li> <li><a name="_5vdc8j2163ef"></a>Consider the true cost of vendor lock-in <ul style="list-style-type: circle;" class="default-list"> <li>Assess long-term <a href="https://www.techtarget.com/searchcloudcomputing/definition/cloud-migration">migration</a> costs before committing to a vendor.</li> <li>Understand how proprietary solutions limit future flexibility.</li> </ul> </li> <li><a name="_2royozww7454"></a>Use a standardized TCO framework <ul style="list-style-type: circle;" class="default-list"> <li>Gartner's TCO model is used for <a href="https://www.techtarget.com/searchdatacenter/definition/infrastructure">IT infrastructure</a> and cloud migration.</li> <li>ISO 15686-5 is used for building lifecycle cost analysis.</li> </ul> </li> </ul> <div class="youtube-iframe-container"> <iframe id="ytplayer-0" src="https://www.youtube.com/embed/FpNVy_nPTbY?autoplay=0&amp;modestbranding=1&amp;rel=0&amp;widget_referrer=null&amp;enablejsapi=1&amp;origin=https://www.techtarget.com" type="text/html" height="360" width="640" frameborder="0"></iframe> </div> <p><em>Not sure what it will cost to run workloads in the cloud? Discover the key variables to <a href="https://www.techtarget.com/searchcloudcomputing/tip/How-to-calculate-your-cloud-TCO">calculate cloud total cost of ownership</a> when comparing on-premises deployment with cloud deployment to avoid costly surprises later on.</em></p> </section> Total cost of ownership (TCO) is an estimation of the expenses associated with purchasing, deploying, managing, using and retiring IT assets, such as a product or piece of equipment. https://cdn.ttgtmedia.com/visuals/digdeeper/5.jpg https://www.techtarget.com/searchdatacenter/definition/TCO Thu, 10 Apr 2025 09:00:00 GMT <p>A northbound interface (NBI) is an application programming interface (<a href="https://www.techtarget.com/searchapparchitecture/definition/application-program-interface-API">API</a>) or <a href="https://www.techtarget.com/searchnetworking/definition/protocol">protocol</a> that allows a lower-level network component to communicate with a higher-level or more central component; conversely, a southbound interface (SBI) allows a higher-level component to send commands to lower-level network components. Northbound and southbound interfaces are most associated with software-defined networking (<a href="https://www.techtarget.com/searchnetworking/definition/software-defined-networking-SDN">SDN</a>), but can also be used in any system that uses a hub-and-spoke -- also called <a href="https://www.techtarget.com/searchnetworking/definition/star-network">star</a> -- or controller-and-nodes architecture.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/sdn-traditional_vs_sdn.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/sdn-traditional_vs_sdn_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/sdn-traditional_vs_sdn_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/sdn-traditional_vs_sdn.png 1280w" alt="Diagram comparing traditional networks and SDN." height="685" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>Software-defined networking simplifies network design and management compared with traditional hardware-oriented networks. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <section class="section main-article-chapter" data-menu-title="Northbound interface and southbound interface in simple terms"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Northbound interface and southbound interface in simple terms</h2> <p>In the context of networking, north and south can be thought of as directions on a map. The north is on the top and south on the bottom of the diagram. Using these cardinal directions helps to visualize internal network communications and how control moves from the top, or north, to the bottom, or south, of a network.</p> <p>All networks consist of both higher-level elements and lower-level elements. Examples of the former include tools for <a href="https://www.techtarget.com/searchnetworking/definition/network-automation">network automation</a>, <a href="https://www.techtarget.com/searchnetworking/definition/network-orchestration">network orchestration</a>, administration consoles, <a href="https://www.techtarget.com/searchnetworking/definition/network-analytics">network analytics</a> and other management systems. Examples of lower-level elements include switches and routers.</p> <p>The higher-level elements control the lower-level ones. At the same time, the network's lower-level components need to communicate with its higher-level components using APIs.</p> <p>Some network designs also have <a href="https://www.techtarget.com/searchnetworking/definition/east-west-traffic">east-west</a> interfaces for communication among peers, also called <a href="https://www.techtarget.com/searchnetworking/definition/peer-to-peer">peer-to-peer</a> communication. These interfaces facilitate communication between the components -- servers, for example -- within the same data center or cloud environment.</p> </section> <section class="section main-article-chapter" data-menu-title="What is northbound and southbound traffic in networking?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What is northbound and southbound traffic in networking?</h2> <p>It is easy to confuse northbound/southbound interface with northbound/southbound data flow or <a href="https://www.techtarget.com/searchnetworking/definition/network-traffic">traffic</a>. The interface -- northbound or southbound -- defines the sender, receiver and data format; it expresses the conceptual level of communication and covers the entire bidirectional API.</p> <p>Conversely, if data or traffic is said to be northbound, southbound or east-west, that merely describes whether it is going toward or away from the network core. It is therefore possible to say that a southbound command went from the core to a node over the northbound interface.</p> <p>Simply put, north-south traffic refers to traffic coming in and going out of the internal network. For example, data traveling between a network <a href="https://www.techtarget.com/whatis/definition/server">server</a> and an external client over the internet is north-south traffic. In contrast, east-west traffic is the internal traffic moving between devices within a network.</p> <p>It's important to understand the flows of both north-south and east-west traffic for security reasons. The more north-south traffic there is in a network, the greater the chances of a <a href="https://www.techtarget.com/searchsecurity/definition/cyber-attack">cyberattack</a> originating from outside that network. <a href="https://www.techtarget.com/whatis/video/An-explanation-of-ransomware">Ransomware</a> attacks and <a href="https://www.techtarget.com/searchsecurity/definition/data-breach">data breaches</a> often occur when there are inadequate security controls to manage and control north-south traffic because these attacks typically originate from outside the network.</p> </section> <section class="section main-article-chapter" data-menu-title="Benefits of northbound and southbound interfaces"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Benefits of northbound and southbound interfaces</h2> <p>Northbound interfaces abstract the inherent complexities of networks, thus simplifying network management. Also, they provide enhanced flexibility and programmability in system configurations, making them essential for efficiently deploying and integrating the various applications and services in the network.</p> <p>Other benefits of NBIs include the following:</p> <ul class="default-list"> <li>Centralized <a href="https://www.techtarget.com/searchnetworking/definition/network-management">network management</a> and control.</li> <li>Efficient <a href="https://www.techtarget.com/searchcio/definition/resource-allocation">resource allocation</a>.</li> <li>Faster application and service deployment.</li> <li>Easier application customization.</li> <li>Real-time monitoring of network infrastructure.</li> <li>Network automation and orchestration.</li> </ul> <p>Northbound interfaces help improve network efficiency, flexibility and agility. By reducing unnecessary complexity, they can also reduce the network's operational expenses and improve its <a href="https://www.techtarget.com/searchcio/definition/ROI">return on investment</a>.</p> <p>Southbound interfaces, on the other hand, play a crucial role in policy implementation. They are also useful for creating actionable and feasible configurations for network devices. Without them, it can be difficult to translate the high-level commands from <a href="https://www.techtarget.com/searchnetworking/definition/network-management-system">network management systems</a> into practical configurations. SBIs can also help to improve network control and management.</p> </section> <section class="section main-article-chapter" data-menu-title="Northbound and southbound interfaces in software-defined networking"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Northbound and southbound interfaces in software-defined networking</h2> <p>In SDN and <a href="https://www.techtarget.com/searchnetworking/What-is-network-virtualization-Everything-you-need-to-know">virtualized networks</a>, the network logical design and data flows are set by software configurations instead of through hardware or physical cabling changes. SDN helps IT administrators to set up highly flexible network infrastructure, which in turn enables organizations to quickly adapt to evolving business requirements. The configuration of the elements is set by the <a href="https://www.techtarget.com/searchnetworking/definition/SDN-controller-software-defined-networking-controller">SDN controller</a>, which sits in the control layer at the center of the network diagram, with north and south established in relation to it.</p> <p>It is important to note that in SDN, the northbound and southbound interfaces are for networking control commands and APIs. The data or traffic carried by the network stays on the data layer and does not traverse the northbound and southbound interfaces.</p> <h3>Northbound interface in SDN</h3> <p>The northbound interface in SDN provides a standardized approach for communication between the highest <a href="https://www.techtarget.com/searchnetworking/definition/Application-layer">application layer</a> and the SDN controller at the middle control layer. It thus enables the efficient exchange of network information between the SDN controller and high-level applications. The application layer consists of network orchestration services, networking design software, operator software or third-party applications that make decisions about the overall structure of the network.</p> <figure class="main-article-image half-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/networking-apis_and_protocols_sdn.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/networking-apis_and_protocols_sdn_half_column_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/networking-apis_and_protocols_sdn_half_column_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/networking-apis_and_protocols_sdn.png 1280w" alt="Diagram of how APIs and protocols work with SDN and SD-WAN." height="334" width="280"> <figcaption> <i class="icon pictures" data-icon="z"></i>Northbound and southbound APIs link network applications and devices to the SDN controller. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <p>In SDN, the operator or orchestration software does not directly issue commands or configurations to the network nodes. Instead, the operator uses the application layer to issue commands to the control layer over the northbound interface.</p> <p>The northbound interface is often a representational state transfer API, or <a href="https://www.techtarget.com/searchapparchitecture/definition/RESTful-API">REST API</a>, exposed by the SDN controller. Various SDN controllers provide northbound interfaces to enable easy <a href="https://www.techtarget.com/searchitoperations/feature/What-does-software-defined-actually-mean">software-defined</a> network management and control. These include the <a target="_blank" href="https://www.opendaylight.org/" rel="noopener">OpenDaylight</a> platform, <a href="https://www.techtarget.com/searchnetworking/definition/ONOS-Open-Network-Operating-System">Open Networking Operating System</a> and Cisco Application Policy Infrastructure Controller.</p> <h3>Southbound interface in SDN</h3> <p>The southbound interface in SDN is the communication between the SDN controller at the middle control layer and the lower networking elements at the data layer. The data layer consists of the physical or virtual network switches and ports.</p> <p>The SDN controller takes the desired state of the network and translates it into specific commands and configurations that are then pushed to the network devices over the southbound interface. SBIs enable the controller to operate efficiently and to make fast adjustments in response to real-time business requirements.</p> <p>Popular southbound interface standards are <a href="https://www.techtarget.com/searchnetworking/definition/SNMP">Simple Network Management Protocol</a>, <a href="https://www.techtarget.com/whatis/definition/OpenFlow">OpenFlow</a> and <a href="https://www.techtarget.com/searchnetworking/definition/OSPF-Open-Shortest-Path-First">Open Shortest Path First</a>. Other protocols available for southbound APIs include <a href="https://www.techtarget.com/searchnetworking/definition/NETCONF">Network Configuration Protocol</a>, <a href="https://www.techtarget.com/searchnetworking/definition/Multiprotocol-Label-Switching-MPLS">Multiprotocol Label Switching</a> and Intermediate System to Intermediate System.</p> <div class="youtube-iframe-container"> <iframe id="ytplayer-0" src="https://www.youtube.com/embed/IRPRX2fVJYY?autoplay=0&amp;modestbranding=1&amp;rel=0&amp;widget_referrer=null&amp;enablejsapi=1&amp;origin=https://www.techtarget.com" type="text/html" height="360" width="640" frameborder="0"></iframe> </div> </section> <section class="section main-article-chapter" data-menu-title="Examples of northbound and southbound interfaces"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Examples of northbound and southbound interfaces</h2> <p>An example of how the northbound and southbound interfaces are used is a network engineer defining a specific data route with network orchestration software. The orchestration software sends the instructions to the SDN controller over the northbound interface. The SDN controller then sends the specific configurations to the physical <a href="https://www.techtarget.com/searchnetworking/definition/switch">switches</a> over the southbound interface.</p> <p>A more detailed example is Microsoft Azure software load balancing. The network controller is at the center layer and runs the <a href="https://www.techtarget.com/searchnetworking/definition/load-balancing">software load balancer</a> (SLB). The network operator sits at the application layer and uses <a href="https://www.techtarget.com/searchwindowsserver/definition/Microsoft-Project-Honolulu">Windows Admin Center</a> to set the desired state. Windows Admin Center uses <a href="https://www.techtarget.com/searchwindowsserver/definition/PowerShell">PowerShell</a> as the northbound interface to send the commands to the SLB. The SLB then sends Border Gateway Protocol (<a href="https://www.techtarget.com/searchnetworking/definition/BGP-Border-Gateway-Protocol">BGP</a>) updates as the southbound interface to the virtual routers on the data layer. If the SLB finds an error in a router, it can automatically send the new configurations to the other routers through the southbound interface BGP, and then send a notification through the northbound interface to alert the operator of the issue.</p> </section> <section class="section main-article-chapter" data-menu-title="Northbound and southbound interfaces in other systems"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Northbound and southbound interfaces in other systems</h2> <p>The concept of northbound and southbound interfaces can also be used in systems with automated control systems and nodes. It can be used when components communicate with different APIs or where an orchestrator is used. The use of separate interfaces contrasts with using a bus architecture.</p> <p><em>In a software-defined network, the data center controllers are essential components. When evaluating controller options, consider factors such as performance and compatibility. Learn what <a href="https://www.techtarget.com/searchdatacenter/tip/Learn-what-SDN-data-center-controllers-do-in-a-network">SDN data center controllers do in a network</a>.</em></p> </section> A northbound interface (NBI) is an application programming interface (API) or protocol that allows a lower-level network component to communicate with a higher-level or more central component; conversely, a southbound interface (SBI) allows a higher-level component to send commands to lower-level network components. https://cdn.ttgtmedia.com/visuals/digdeeper/6.jpg https://www.techtarget.com/whatis/definition/northbound-interface-southbound-interface Wed, 02 Apr 2025 14:00:00 GMT <p>A green data center is a repository for the storage, processing, management and dissemination of data in which the physical space and the mechanical and electrical subsystems are designed to maximize energy efficiency and minimize the environmental impact. The construction and operation of a green data center includes advanced technologies and strategies.</p> <p>The following are examples of some of the technologies and strategies used in green data center initiatives:</p> <ul class="default-list"> <li>Minimized building footprints with physical spaces optimized for effective airflow.</li> <li>Low-emission building materials, carpets and paints.</li> <li>Sustainable landscaping.</li> <li>Extensive use of virtualization to maximize hardware utilization while reducing space and heat generation.</li> <li>Use of AI platforms to analyze data center operations and optimize data center infrastructure management (DCIM) techniques.</li> <li><a href="https://www.techtarget.com/sustainability/definition/e-waste">E-waste</a> recycling.</li> <li>Catalytic converters on backup generators.</li> <li>Alternative cooling technologies, including heat pumps, <a href="https://www.techtarget.com/searchdatacenter/definition/data-center-evaporative-cooling-swamp-cooling">evaporative cooling</a>, natural cooled facilities located in cold geographic locations and liquid immersion cooling for servers and other computing equipment.</li> <li>Heat recovery and redirection designs that redirect and reuse heat from data center facilities to other cooler areas of the building.</li> <li>Alternative and renewable energy sources, such as photovoltaic technology, biofuels and wind technology.</li> <li>Hybrid and electric vehicles.</li> </ul> <p>Building and certifying a green data center or other facility can be expensive upfront, but long-term cost savings can be realized on operations and maintenance. Green facilities offer employees a healthy, comfortable work environment and enhance relations with local communities.</p> <p>Environmentalists and, increasingly, the general public are pressuring governments to offer green incentives. Companies sometimes receive tax incentives and other monetary support for the development and use of <a href="https://www.techtarget.com/searchdatacenter/tip/Considerations-for-sustainable-data-center-design">environmentally responsible technologies</a>.</p> <div class="youtube-iframe-container"> <iframe id="ytplayer-0" src="https://www.youtube.com/embed/yGkfBo2iSiI?si=_iJvhdBHlh6lhAPt?autoplay=0&amp;modestbranding=1&amp;rel=0&amp;widget_referrer=null&amp;enablejsapi=1&amp;origin=https://www.techtarget.com" type="text/html" height="360" width="640" frameborder="0"></iframe> </div> <section class="section main-article-chapter" data-menu-title="Why do you need a green data center?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Why do you need a green data center?</h2> <p>Green data centers have become essential for enterprise computing. A properly designed and well-implemented green data center design can use a range of energy-efficient technologies to do the following:</p> <ul class="default-list"> <li>Lower power consumption, which is one of the largest data center expenses.</li> <li>Reduce carbon emissions, which demonstrates environmental responsibility, improves public brand perception and helps the environment.</li> <li>Conserve valuable resources, such as water and fossil fuels.</li> <li>Improve the sustainability of data center operations, which is a critical concern for <a href="https://www.techtarget.com/searchdisasterrecovery/definition/business-continuity">business continuity</a>.</li> </ul> <p>Establishing a green data center can be complex and expensive. Traditional data centers can be upgraded or retrofitted with green technologies, such as virtualization, AI-driven DCIM, and refurbishing and recycling programs. However, the deployment of a full-service green data center with a comprehensive suite of green technologies and strategies typically requires the construction of a new data center facility.</p> </section> <section class="section main-article-chapter" data-menu-title="What are the benefits of a green data center?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What are the benefits of a green data center?</h2> <p>Adopting a green approach to data center energy and environment management is a significant investment. However, over time, these data centers provide a range of advantages. Some benefits of green data centers are the following:</p> <ul class="default-list"> <li>Lower long-term operating costs.</li> <li>Reduced physical space requirements.</li> <li>Less carbon emissions and a smaller <a href="https://www.techtarget.com/whatis/definition/carbon-footprint">carbon footprint</a> due to fewer and more energy-efficient data center gear.</li> <li>Decreased water use, expanding the range of suitable data center locations while reducing contention for water with local communities.</li> <li>Reduced waste output from packaging reductions, gear redeployments and recycling.</li> <li>Enhanced business continuity and regulatory compliance postures.</li> <li>Reduced electricity consumption.</li> <li>More emphasis on renewable and sustainable data center technology and resources.</li> </ul> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/actions_to_help_reduce_ewaste-f.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/actions_to_help_reduce_ewaste-f_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/actions_to_help_reduce_ewaste-f_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/actions_to_help_reduce_ewaste-f.png 1280w" alt="list of ways to reduce e-waste" height="413" width="559"> <figcaption> <i class="icon pictures" data-icon="z"></i>Four ways data center operators can deal with e-waste. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> </section> <section class="section main-article-chapter" data-menu-title="Green data center performance metrics"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Green data center performance metrics</h2> <p>Numerous metrics have been developed to measure energy use and sustainability. They help demonstrate and certify that buildings are using energy efficiently and don't harm the environment. Organizations typically adopt and track several green energy metrics to provide the most complete picture of energy efficiency and sustainable practices and operation.</p> <ul class="default-list"> <li><b>Power usage effectiveness. </b>Developed in 1997 by The Green Grid, part of the Information Technology Industry Council, <a href="https://www.techtarget.com/searchdatacenter/definition/power-usage-effectiveness-PUE">PUE</a> measures the power consumption of a data center. It's the ratio of the total power provided to the data center divided by the power the IT equipment in the data center uses. The goal is to have the ratio come as close as possible to one. A ratio of one means all of the power provided to the data center is being used by the computing equipment, effectively eliminating heat and waste. Ratios greater than one indicate lower efficiency.</li> <li><b>Data center Infrastructure efficiency. </b><a href="https://www.techtarget.com/searchdatacenter/definition/data-center-infrastructure-efficiency-DCIE">DCiE</a> measures how effectively a data center uses the available infrastructure. It divides the power consumed by IT gear by the total power the data center uses. DCiE is the inverse of PUE. A DCiE approaching one indicates greater energy efficiency. As with PUE, DCiE is a direct gauge of efficiency and is often used to help engineers measure the impact of infrastructure changes on efficiency.</li> <li><b>Carbon usage effectiveness.</b> The Green Grid also developed<b> </b>the<b> </b>CUE metric to show if a data center has attained its sustainability goals. It's the ratio of carbon dioxide emissions the data center generates divided by the energy consumption of data center equipment. The goal is to have the lowest possible value, which indicates that the data center is effectively controlling its carbon dioxide emissions and carbon footprint.</li> <li><b>Water usage effectiveness. </b>The WUE metric evaluates a data center's water consumption relative to its energy output. It's determined by dividing the total annual water use in liters by the total annual energy use in kilowatt-hours. The result is WUE expressed in liters per <a href="https://www.techtarget.com/whatis/definition/watt-hour-Wh">kWh</a>. WUE is most useful for facilities that rely on water-based cooling systems.</li> <li><b>Renewable energy usage.</b> REU is the ratio of energy derived from renewable sources, including photovoltaic, wind and hydroelectric power, divided by the total electricity used by the data center. This renders a ratio, but REU is typically expressed as a percentage. When the REU reaches 100%, all of the facility's energy use is from renewable sources. REU is most helpful for green data centers that emphasize sustainability through renewable energy sources.</li> <li><b>Energy reuse factor.</b> ERF is a more exotic metric that expresses energy efficiency in terms of re-use. ERF is a ratio of the energy re-used, such as using waste heat from servers to heat other spaces, versus the total energy the data center uses. This ratio is expressed as a percentage, and is typically low, but larger percentages indicate more effective energy reuse.</li> <li><b>Data center carbon footprint.</b> The DCCF metric gauges the environmental impact of a data center. It's determined by multiplying the total power the data center consumes by the carbon emission factor of the power source. A lower DCCF means a lower carbon impact. Engineers often use this metric to determine the sustainability of the data center and as justification for investments in renewable, low-carbon emission energy sources.</li> </ul> <figure class="main-article-image full-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/data_center-pue.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/data_center-pue_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/data_center-pue_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/data_center-pue.png 1280w" alt="How power usage effectiveness (PUE) works" height="286" width="560"> <figcaption> <i class="icon pictures" data-icon="z"></i>Power usage effectiveness is a metric used to assess the efficiency of a data center. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> </section> <section class="section main-article-chapter" data-menu-title="Energy efficiency certifications"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Energy efficiency certifications</h2> <p>Three certifications are available to validate that a building or IT device is energy-efficient and environmentally friendly:</p> <ol type="1" start="1" class="default-list"> <li><b>Leadership in Energy and Environmental Design.</b> The U.S. Green Building Council created this certification. <a href="https://www.techtarget.com/searchdatacenter/definition/LEED-Leadership-in-Energy-and-Environmental-Design">LEED</a> certification of a building means it has satisfied a rigorous set of criteria to reduce energy consumption and be environmentally friendly.</li> <li><b>Energy Star.</b> Developed by the U.S. Environmental Protection Agency and the U.S. Department of Energy, the <a href="https://www.techtarget.com/searchdatacenter/definition/Energy-Star">Energy Star</a><b> </b>designation certifies that a machine or device is energy-efficient. Use of Energy Star-certified products have <a target="_blank" href="https://www.energystar.gov/about?s=mega" rel="noopener">saved residential and businesses users more than 5 trillion kilowatts</a> of electricity since 1992.</li> <li><b>Certified Energy Efficient Datacenter Award.</b> CEEDA is a certification framework designed around a mix of standards, including ones from the American Society of Heating, Refrigerating, and Air Conditioning Engineers; Energy Star; the European Code of Conduct; European Telecommunications Standards Institute; Green Grid metrics; and the International Organization for Standardization. This global certification program independently recognizes the successful implementation of energy-efficiency practices in data centers across a variety of specialized disciplines.</li> </ol> </section> <section class="section main-article-chapter" data-menu-title="Components of a green data center"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Components of a green data center</h2> <p>Virtually every component of a data center, from how the building is constructed to the equipment in use, can be made more energy-efficient and environmentally friendly.</p> <figure class="main-article-image half-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/storage-energy_efficient_storage_checklist-h.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/storage-energy_efficient_storage_checklist-h_half_column_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/storage-energy_efficient_storage_checklist-h_half_column_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/storage-energy_efficient_storage_checklist-h.png 1280w" alt="list of factors affecting energy-efficiency of storage" height="427" width="279"> <figcaption> <i class="icon pictures" data-icon="z"></i>Efficient storage technology is a key part of a green data center. See what it takes to have energy-efficient storage. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <p>Energy efficiency and environmental considerations are essential design components when upgrading an existing data center or building a new one. Organizations planning a green data center can use a design firm with experience in designing energy-efficient and environmentally friendly buildings.</p> <p>Design considerations and components of a green data center include the following:</p> <ul class="default-list"> <li><b>Cold and hot aisles.</b> Data center servers are placed and contained in <a href="https://www.techtarget.com/searchdatacenter/definition/hot-cold-aisle">cold and hot aisles</a> that enable hot air to be pumped to air conditioner returns and cold air from cold aisles to where it's needed for cooling.</li> <li><b>Free air cooling.</b> <a href="https://www.techtarget.com/searchdatacenter/definition/free-cooling">Free cooling</a> systems use outdoor air to cool data centers that are strategically located in cooler climates.</li> <li><b>Liquid cooling.</b> Immersion cooling technologies submerge data center gear in a circulated bath of nonconductive oil that contains and carries heat more efficiently than air cooling while using less power in chiller and circulation equipment than traditional heating, ventilation and air conditioning gear.</li> <li><b>Low-power servers.</b> These servers work well in data centers. Their low energy consumption makes them more efficient than traditional servers.</li> <li><b>Virtualized servers.</b> Virtualization lets a single physical server host multiple virtual server instances. One physical server can function as many different servers and reduce the total server count in the data center.</li> <li><b>Modular data centers.</b> These energy-efficient data centers are portable and can be quickly set up wherever they're needed. They're also called <i>data centers in a box</i>.</li> <li><b>Evaporative cooling.</b> Various technologies, such as evaporation pads and high-pressure spray systems, reduce heat through the evaporation of water.</li> <li><b>Heat recovery and reuse.</b> Waste heat from data center power use is reused to heat other facilities.</li> <li><b>Ultrasonic humidification.</b> Energy-efficient ultrasound is used to create the moisture needed to establish proper environmental conditions to run some devices in a data center. For example, adequate humidity reduces static accumulation and potentially damaging static discharges.</li> <li><b>Renewable energy use.</b> Green data centers will typically integrate one or more renewable energy sources, such as photovoltaic, wind, hydroelectric or biofuel installations. This reduces their carbon footprint, improves sustainability and reduces dependence on traditional utilities.</li> <li><b>Infrastructure monitoring and management.</b> Software technologies such as DCIM let data center operators monitor energy use, optimize resource allocation, ensure safe operational environments, and oversee facility and IT infrastructure performance.</li> <li><b>Building design.</b> Factors such as building size and positioning, insulation type, natural lighting and air handling can reduce ongoing energy demands and affect data center efficiency.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="Cloud computing and green data centers"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Cloud computing and green data centers</h2> <p>The global rise of cloud computing has altered the calculus of green data center design and implementation for many organizations. In effect, cloud computing allows an organization to use computing resources and services remotely and on demand. This treats computing as if it were a utility such as electricity or water.</p> <p>For most organizations, the availability of cloud computing has reduced, and even eliminated, the demand for green data center design. Cloud computing mitigates issues such as hardware counts, power supply and use, cooling needs and physical space.</p> <p>For example, an organization might opt to migrate key workloads and data to a cloud provider rather than invest the time, financial capital and personnel in building and operating a green data center. Some of the most progressive businesses might forego a data center entirely and commit to a cloud-first or cloud-native posture for their business software and services.</p> <p>However, for cloud computing providers with many large data centers around the world, the need to design, build and manage green data centers is more acute than ever.</p> <p><i>Learn more about what factors to consider to </i><a href="https://www.techtarget.com/searchdatacenter/feature/Get-started-with-green-energy-for-your-data-center"><i>run a sustainable data center</i></a><i>.</i></p> </section> A green data center is a repository for the storage, processing, management and dissemination of data in which the physical space and the mechanical and electrical subsystems are designed to maximize energy efficiency and minimize the environmental impact. https://cdn.ttgtmedia.com/visuals/digdeeper/5.jpg https://www.techtarget.com/searchdatacenter/definition/green-data-center Thu, 06 Mar 2025 09:00:00 GMT <p>Data center infrastructure management (DCIM) is the convergence of <a href="https://www.techtarget.com/searchdatacenter/definition/infrastructure">IT infrastructure</a>, operations and building facilities functions within an organization. A DCIM initiative aims to provide administrators with holistic aspects of a data center's performance so that energy, equipment and physical data center floor space are used as efficiently as possible.</p> <p>DCIM software measures, monitors and manages the IT equipment and supporting infrastructure of data centers. This enables data center operators to run efficient operations while improving infrastructure design planning. DCIM software can be hosted on-premises or in the cloud as software as a service.</p> <p>Data center infrastructure management tools can help administrators locate and identify relationships between a building and its IT systems. For example, energy-monitoring <a href="https://www.techtarget.com/whatis/definition/sensor">sensors</a> and supporting hardware can be installed to measure and manage the energy consumption of IT equipment and infrastructure components such as power distribution units (<a href="https://www.techtarget.com/searchdatacenter/definition/power-distribution-unit-PDU">PDUs</a>). The same system can then analyze <a href="https://www.techtarget.com/searchdatacenter/definition/power-usage-effectiveness-PUE">power usage effectiveness</a> and cooling system energy efficiency.</p> <section class="section main-article-chapter" data-menu-title="How does DCIM work?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How does DCIM work?</h2> <p>DCIM tools collect and centralize an assortment of essential data derived from infrastructure components deployed across a data center -- and often broader devices such as <a href="https://www.techtarget.com/whatis/definition/endpoint-device">endpoint systems</a> -- along with an array of data center power, temperature and other environmental parameters. This data is stored and processed against an established baseline to provide historical, real-time predictive insights about the organization's infrastructure. When properly implemented, DCIM can offer comprehensive oversight and management of the entire infrastructure through a single unified interface or <a href="https://www.techtarget.com/searchdatacenter/definition/single-pane-of-glass">single pane of glass</a>.</p> <p>DCIM components typically involve several key elements, including the following:</p> <ul class="default-list"> <li><b>A centralized database.</b> This database contains details about every managed system or device. This is a detailed exercise in asset management, which can include servers, storage and network equipment, power distribution units, systems such as heating, ventilation and air conditioning, and environmental sensors such as temperature and humidity.</li> <li><b>DCIM software.</b> This software accesses the centralized database and collects and stores desired performance metrics for each managed device. The software can display metrics such as power consumption and energy efficiency in real time and report details such as system use over time. Historical data can indicate trends vital for data center tasks such as <a href="https://www.techtarget.com/searchdatacenter/definition/data-center-capacity-planning">capacity planning</a> or recommend infrastructure optimizations.</li> <li><b>A dashboard user interface (UI).</b> Real-time DCIM software shares much of its data through a real-time dashboard UI and can also generate real-time alerts, such as a problem with system health or an unexpected temperature increase in a server rack, along with periodic performance reporting over time. The UI can typically produce detailed visualizations such as charts and graphs.</li> <li><b>Software connectors.</b> Although some systems and devices can communicate with DCIM software directly, software connectors such as <a href="https://www.techtarget.com/searchapparchitecture/definition/application-program-interface-API">application programming interfaces</a> are often needed to include equipment in the DCIM scope and collect desired data. Connectors can be obtained from DCIM or system vendors or sometimes created in-house using standard programming tools and techniques. For DCIM to be effective, DCIM software must recognize and interoperate with all systems and devices under management.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="Why is DCIM important?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Why is DCIM important?</h2> <p>DCIM offers various benefits, but the overarching importance is its <i>scope</i>. IT has long suffered from disconnected -- and sometimes competing -- considerations. For example, different groups might manage different servers or storage; an entirely different team might be responsible for the facility; IT and business managers would need to determine what's in service and how well it's working. Each concern might employ specialized or vendor-specific management tools to address its areas of responsibility.</p> <p>These types of disconnects made IT practices slow and inefficient. It was almost impossible for business and technology leaders to attain a comprehensive view of all gear and its operational behaviors. Gaps in insight and control wasted time, increased operational costs, made IT inflexible and difficult to troubleshoot and increasingly left the business vulnerable to compliance and business continuity problems.</p> <p>In the most direct terms, DCIM fixes the following disconnects by centralizing knowledge and control:</p> <ul class="default-list"> <li>DCIM builds and maintains a comprehensive record of all equipment, devices and services in operation.</li> <li>DCIM compiles a clear set of dependencies and interrelationships among equipment, devices, facilities and services, including group- or department-level ownership.</li> <li>DCIM can manage data centers through a single interface.</li> <li>DCIM enables real-time observability, alerting and comprehensive reporting that supports strong change management and accelerates troubleshooting.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="What are the benefits of DCIM?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What are the benefits of DCIM?</h2> <p>Benefits of DCIM include the following:</p> <ul class="default-list"> <li><b>Increased uptime.</b> Monitoring equipment and constructing reports on gathered data help an organization determine its data center health and whether equipment needs replacing.</li> <li><b>Better resource use.</b> DCIM enables data center managers to oversee the health and performance of physical resources available across the enterprise and facility resources such as power usage, cooling systems and physical space. Companies commonly use this insight to optimize IT resource efficiency.</li> <li><b>Capacity management.</b> DCIM software features help organizations model and allocate space for new hardware and equipment and manage their power chain.</li> <li><b>Proactive incident management.</b> Real-time management features give facility managers insight into the state of their data centers now and into the future.</li> <li><b>Energy management.</b> DCIM enables organizations to measure their energy consumption for all IT-related equipment. This insight helps them reduce energy usage and costs.</li> <li><b>Improved productivity.</b> DCIM enables IT departments to monitor multiple locations remotely. This benefits edge data centers where automated alerts can help IT track physical and logical components and warn of component failures.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="What are the challenges of DCIM?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What are the challenges of DCIM?</h2> <p>DCIM also has the following disadvantages:</p> <ul class="default-list"> <li><b>Security.</b> DCIM connects multiple critical systems, increasing the number of potential entry points.</li> <li><b>Siloed data.</b> Frequent problems can occur when integrating data from diverse sources like power systems, cooling systems and IT equipment. Data might not be compatible, require transformations or be missing, potentially leading to gaps in management.</li> <li><b>Environmental monitoring.</b> Servers and other IT equipment in racks should be carefully monitored with temperature sensors to ensure no equipment overheats. Depending on the environment, this could be expensive.</li> <li><b>Facility systems.</b> Older data centers might lack a system or not have updated equipment. For example, a data center might not have real-time utility energy metering. It might have to base total energy usage on its utility bill, which could cost more in terms of power consumption. Likewise, the lack of temperature sensors in older equipment could mean the organization must invest in newer equipment.</li> <li><b>Skill sets.</b> DCIM systems can be cumbersome to implement and often demand discipline and consistent practice to manage properly. Staff must be trained and experienced with the selected DCIM software and follow carefully vetted DCIM practices. This requires skill with DCIM tools and business processes.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="What are the components of DCIM?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What are the components of DCIM?</h2> <p>DCIM components surround the architecture and support IT functions and infrastructure. These specific components and capabilities include the following:</p> <figure class="main-article-image half-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/6_components_of_a_dcim_architecture-h.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/6_components_of_a_dcim_architecture-h_half_column_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/6_components_of_a_dcim_architecture-h_half_column_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/6_components_of_a_dcim_architecture-h.png 1280w" alt=": A chart listing the main components of a DCIM architecture. " height="250" width="279"> <figcaption> <i class="icon pictures" data-icon="z"></i>DCIM components support an organization's IT functions and infrastructure. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <ul class="default-list"> <li><b>Physical architecture.</b> This includes floor space consideration for hardware components such as servers, network switches, airflow, cooling systems and PDUs.</li> <li><b>Materials catalog.</b> DCIM technologies contain libraries of material lists. These could include basic specifications, servers, storage and networking equipment or PDUs. The vendors update and maintain these libraries regularly.</li> <li><b>Capacity planning.</b> Planning for future capacity includes estimating space, computer hardware, software and connection infrastructure resources. DCIM software can construct models for future scenarios based on specified limitations.</li> <li><b>Change management.</b> DCIM must account for hardware replacement to avoid risks like downtime or malfunctions. It must also account for process changes and can be used to maintain an audit trail of requests and work orders.</li> <li><b>Software integration.</b> DCIM software can integrate with existing management software to better track automated and integrated data center workflows.</li> <li><b>Data analysis.</b> DCIM software also collects real-time data, such as hardware metrics, which helps mitigate outages or other incidents.</li> </ul> </section> <section class="section main-article-chapter" data-menu-title="How to get started with DCIM"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How to get started with DCIM</h2> <p>A sound strategy is for an organization to implement DCIM in phases, starting with the bare minimum and upgrading functionality in small steps. Implementing DCIM this way makes the process of migrating less complex. For example, an organization can move from creating spreadsheets to implementing automated data center asset management software. From these quick and easy wins, the business can work on adopting other DCIM aspects such as capacity, power and environment.</p> <p>Organizations should <a target="_blank" href="https://www.gartner.com/reviews/market/data-center-infrastructure-management-tools" rel="noopener">look for DCIM software</a> that fits their physical architecture and needs for monitoring, change management, scaling and capacity planning, material cataloging and <a href="https://www.techtarget.com/searchbusinessanalytics/definition/data-visualization">data visualizations</a>. Data visualizations are how data center information is presented, such as in a dashboard, trend charts, reports, rack and row elevations and floor layout plans.</p> <div class="youtube-iframe-container"> <iframe id="ytplayer-0" src="https://www.youtube.com/embed/WpaM7z9TPuo?autoplay=0&amp;modestbranding=1&amp;rel=0&amp;widget_referrer=null&amp;enablejsapi=1&amp;origin=https://www.techtarget.com" type="text/html" height="360" width="640" frameborder="0"></iframe> </div> <p>DCIM tools can be large, complex platforms. Once adopted and implemented, they can be incredibly difficult to change, making the product vetting and selection process critical. DCIM proof-of-concept initiatives are highly recommended to ensure that the management tool being considered is a proper fit for business needs, offers necessary integrations, presents a manageable learning curve, and offers future capabilities and scalability for the business.</p> <p><i>The increasing carbon footprint of artificial intelligence is alarming. Discover </i><a href="https://www.techtarget.com/sustainability/feature/How-CIOs-can-reduce-AIs-negative-environmental-impact"><i>how leaders can promote energy efficiency and sustainability for AI</i></a><i>.</i></p> </section> Data center infrastructure management (DCIM) is the convergence of IT infrastructure, operations and building facilities functions within an organization. https://cdn.ttgtmedia.com/visuals/digdeeper/6.jpg https://www.techtarget.com/searchdatacenter/definition/data-center-infrastructure-management-DCIM Tue, 04 Mar 2025 12:15:00 GMT <p>ISO 50001 (International Organization for Standardization 50001) is a voluntary standard for designing, implementing and maintaining an energy management system. ISO 50001 was developed by an <a href="https://www.techtarget.com/searchdatacenter/definition/ISO">ISO</a> technical committee and originally published in 2011. It was updated in 2018 and 2024 to align with other ISO standards and climate action changes.</p> <p>ISO 50001, which uses a vendor- and technology-agnostic approach to enterprise energy management, was developed in response to a request from the United Nations Industrial Development Organization. The standard is intended to <a href="https://www.techtarget.com/searchcio/feature/Where-ISO-certification-fits-in-a-risk-mitigation-strategy">help organizations develop and implement</a> an energy management system.</p> <p>According to the ISO website, an energy management system is a framework for implementing technical and management strategies that will significantly cut energy costs and <a href="https://www.techtarget.com/whatis/definition/greenhouse-gas">greenhouse gas emissions</a> over time. System components include the creation of an energy policy, objectives for improving the efficient use of energy, a timeline with target dates for meeting objectives and an action plan that specifies exactly how the organization's objectives will be met.</p> <p>ISO 50001 uses the <a href="https://www.techtarget.com/whatis/definition/PDCA-plan-do-check-act">plan-do-check-act</a> framework, which is also supported by the U.S. Department of Energy. During the <i>planning phase, </i>the organization sets objectives and targets, using current energy efficiency measurements to establish a baseline. During the <i>do phase</i>, the organization implements actions to <a href="https://www.techtarget.com/searchdatacenter/tip/Four-ways-to-reduce-data-center-power-consumption">improve energy efficiency</a>. During the <i>check phase</i>, the organization measures and evaluates its energy performance and compares the results <a href="https://www.techtarget.com/searchdatacenter/tip/How-much-energy-do-data-centers-consume">to its baseline</a>. During the <i>act phase</i>, the organization decides what changes to make to improve energy performance. The cycle then continuously repeats with a new planning phase.</p> <p>With continuous improvement, ISO 50001 will help an organization <a href="https://www.techtarget.com/searchdatacenter/tip/Use-ISO-500012018-as-a-guide-for-green-data-centers">make better use of its existing energy-consuming assets</a>, create transparency and improve communication about energy consumption, promote energy management best practices and help prioritize the implementation of energy-efficient technology.</p> <p>The ISO does not issue certifications for ISO 50001 compliance, but an accredited third-party agency can grant certifications. It is recommended that organizations interested in acquiring ISO 50001 certification contact and interview several accredited agencies before committing to one. A list of accredited evaluators is available on the ISO website. The ISO 50001 certification is valid for three years, and a mandatory assessment is done every year.</p> ISO 50001 (International Organization for Standardization 50001) is a voluntary standard for designing, implementing and maintaining an energy management system. https://cdn.ttgtmedia.com/visuals/digdeeper/1.jpg https://www.techtarget.com/searchdatacenter/definition/ISO-50001-International-Organization-for-Standardization-50001 Fri, 24 Jan 2025 09:00:00 GMT <p>A hot spot/cold spot is an undesirable, tightly focused local temperature variation that often occurs when data center equipment is improperly cooled.</p> <p>Hot spots are a serious issue that can result in equipment damage. Consider a typical <a href="https://www.techtarget.com/searchdatacenter/definition/blade-server">blade server</a>, which might individually include several processors and a large quantity of memory. These components can <a href="https://www.techtarget.com/searchdatacenter/tip/How-much-energy-do-data-centers-consume">demand a significant amount of power</a> -- much of that power is dissipated as <a href="https://www.techtarget.com/whatis/definition/heat">heat</a>. When many blade servers are installed into a tight, highly integrated chassis, the <a href="https://www.techtarget.com/searchdatacenter/tip/Evaluate-air-cooling-vs-liquid-cooling-for-your-data-center">available cooling resources</a> might dissipate the cumulative heat adequately. This results in one or more areas of excess temperature within the blade chassis. In extreme cases, the excess heating might cause processors, memory modules or other blade server components to fail.</p> <p><a href="https://www.techtarget.com/searchdatacenter/tip/Data-center-cooling-systems-and-technologies-and-how-they-work">Proper cooling design</a> is critical in addressing the hot spots encountered with modern, dense IT equipment. When a hot spot is discovered, it might be necessary to reconfigure or relocate the affected equipment or provide supplemental local cooling (point cooling) to address the condition.</p> <div class="youtube-iframe-container"> <iframe id="ytplayer-0" src="https://www.youtube.com/embed/8fjdpJ-9JI4?autoplay=0&amp;modestbranding=1&amp;rel=0&amp;widget_referrer=null&amp;enablejsapi=1&amp;origin=https://www.techtarget.com" type="text/html" height="360" width="640" frameborder="0"></iframe> </div> <p>A cold spot might develop on or around equipment that receives excess cooling -- perhaps the equipment is located too close to a floor grate or inlet. When a cold spot is discovered, it's important to <a href="https://www.techtarget.com/searchdatacenter/tip/Tips-to-build-a-data-center-airflow-management-strategy">review the airflow</a> and rectify any disruptions. When a surface becomes unusually cold, it often means that the cooling airflow is being obstructed. Cold spots might also pose a moisture danger. If a surface temperature <a href="https://www.techtarget.com/searchdatacenter/tip/Data-center-temperature-and-humidity-guidelines">falls below the surrounding air's dew point</a>, condensation may form and allow liquid water to accumulate in a location that might damage equipment.</p> A hot spot/cold spot is an undesirable, tightly focused local temperature variation that often occurs when data center equipment is improperly cooled. https://cdn.ttgtmedia.com/visuals/digdeeper/1.jpg https://www.techtarget.com/searchdatacenter/definition/Hot-spot-cold-spot Wed, 08 Jan 2025 09:00:00 GMT <p>A diode is a specialized electronic component, specifically a two-terminal semiconductor device, with two electrodes called the <a href="https://www.techtarget.com/whatis/definition/anode">anode</a> and the <a href="https://www.techtarget.com/whatis/definition/cathode">cathode</a>. It has several uses: as a one-way switch that enables current to flow easily in one direction while restricting its flow in the other direction, as a converter that changes alternating current into pulsating direct current, and as a circuit protector that limits the voltage in the circuit. Diodes are also used in clamping circuits and logical gates.</p> <section class="section main-article-chapter" data-menu-title="Diode construction"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Diode construction</h2> <p>Most diodes are made with <a href="https://www.techtarget.com/whatis/definition/semiconductor">semiconductor</a> materials such as silicon, germanium or selenium. Other semiconducting materials like <a href="https://www.techtarget.com/whatis/definition/gallium-arsenide-GaAs">gallium arsenide</a> can also be used to construct diodes.</p> <div class="youtube-iframe-container"> <iframe id="ytplayer-0" src="https://www.youtube.com/embed/OiV8WhI5gpw?autoplay=0&amp;modestbranding=1&amp;rel=0&amp;widget_referrer=null&amp;enablejsapi=1&amp;origin=https://www.techtarget.com" type="text/html" height="360" width="640" frameborder="0"></iframe> </div> <p>Some diodes are composed of metal electrodes in a chamber evacuated or filled with a pure elemental gas at low pressure. Two equivalently doped P-type and N-type semiconductors are joined together, and electrode terminals are attached to each end to create a P-N junction diode. This results in the holes in the P-type semiconductor attracting <a href="https://www.techtarget.com/whatis/definition/electron">electrons</a> from the N-type material. It is this attraction that causes the electrons to diffuse and occupy the holes, and lets current pass through the diode.</p> <p>The P-type material with its excess holes is positively charged. It is known as the positive lead or anode. In contrast, the N-type material with excess electrons is negatively charged and known as the negative lead or cathode.</p> <figure class="main-article-image half-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/the_components_of_a_diode-h.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/the_components_of_a_diode-h_half_column_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/the_components_of_a_diode-h_half_column_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/the_components_of_a_diode-h.png 1280w" alt="Diagram of a diode." height="155" width="279"> <figcaption> <i class="icon pictures" data-icon="z"></i>Example of a diode with its two electrodes called the anode and the cathode. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <p>The movement of electrons also creates what is known as a depletion layer between the P-type and N-type materials. The depletion layer, also known as a depletion region, is a thin area that's depleted of charge due to the movement, or diffusion, of electrons from the N-type material to the P-type material, and the diffusion of holes from the P-type material to the N-type material. Due to the charge depletion, this layer offers high resistance, meaning it opposes the further flow of electrons from the N-side to the P-side.</p> </section> <section class="section main-article-chapter" data-menu-title="What is the function of a diode?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What is the function of a diode?</h2> <p>The fundamental property of a diode is to conduct electric <a href="https://www.techtarget.com/whatis/definition/current">current</a> in only one direction. Diodes can be used as rectifiers, signal limiters, voltage regulators, switches, signal modulators, signal mixers, signal demodulators and <a href="https://www.techtarget.com/whatis/definition/oscillator">oscillators</a>.</p> <p>Semiconductor diodes can be designed to produce direct current when visible light, infrared transmission or ultraviolet energy strikes them. These diodes are known as photovoltaic cells and are the basis for <a href="https://www.computerweekly.com/feature/Datacentres-energy-efficiency-investment-priorities-in-the-race-to-net-zero">solar electric energy</a> systems and photosensors.</p> <p>Yet another form of diode, commonly used in electronic and computer equipment, emits visible light or infrared energy when current passes through it. Such a device is the familiar light-emitting diode (LED).</p> </section> <section class="section main-article-chapter" data-menu-title="How does a diode work?"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How does a diode work?</h2> <p>When the cathode is negatively charged relative to the anode at a voltage greater than a certain minimum, called forward breakover, then current flows through the diode. If the cathode is positive with respect to the anode, is at the same voltage as the anode or is negative by an amount less than the forward breakover voltage, then the diode does not conduct current. This is a simplistic view, but is true for diodes operating as rectifiers, switches and limiters. The forward breakover voltage is approximately 0.6 V for silicon devices, 0.3 V for germanium devices and 1 V for selenium devices.</p> <p>The above general rule notwithstanding, if the cathode voltage is positive relative to the anode voltage by a great enough amount, the diode will conduct current. The voltage required to produce this phenomenon, known as the avalanche voltage, varies greatly depending on the nature of the <a href="https://www.techtarget.com/searchdatacenter/tip/A-primer-on-AI-chip-design">semiconductor material</a> from which the device is fabricated. The avalanche voltage can range from a few volts up to several hundred volts.</p> <p>When an analog signal passes through a diode operating at or near its forward breakover point, the signal waveform is distorted. This nonlinearity enables <a href="https://www.techtarget.com/searchnetworking/definition/modulation">modulation</a>, demodulation and signal mixing. In addition, signals are generated at <a href="https://www.techtarget.com/whatis/definition/harmonic">harmonics</a>, or integral multiples of the input frequency. Some diodes also have a characteristic that is imprecisely termed <i>negative resistance</i>. Diodes of this type, with the application of a voltage at the correct level and the polarity, generate analog signals at microwave radio frequencies.</p> </section> <section class="section main-article-chapter" data-menu-title="Forward bias and reverse bias in a diode"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Forward bias and reverse bias in a diode</h2> <p>A semiconductor diode is forward biased when the anode's voltage is more positive than the cathode's voltage, causing the diode to conduct current easily. In contrast, it is reverse biased when the cathode voltage is more positive than the anode voltage, increasing <a href="https://www.techtarget.com/whatis/definition/resistance">resistance</a> and hindering the current's flow.</p> <h3>Forward bias</h3> <p>In a semiconductor diode, forward bias occurs when the P-region connects with the positive terminal of a battery while the N-region is connected with its negative terminal. The voltage is applied in the forward direction, resulting in more holes getting pumped into the P-region and more electrons getting pumped into the N-region. The holes and electrons moving in opposite directions put pressure on the depletion layer, causing it to shrink and the diode to offer less resistance.</p> <p>The electrons in the N-region overcome the depletion layer and flow to the P-region, causing the anode to be more positively biased than the cathode, which results in the current flow increasing in one direction. This process is known as rectification. The voltage level at which the depletion region completely disappears and causes the current to flow most easily -- it increases exponentially to the applied voltage -- is known as the threshold voltage, cutoff voltage or firing voltage.</p> <h3>Reverse bias</h3> <p>Reverse bias occurs when the diode's P-type region is connected to the battery's negative terminal while the N-type is connected to the positive terminal. This configuration causes the P-type to lose holes and the N-type to lose electrons. These actions cause electrons to get pulled away from the P-N junction and the diode to be depleted of charge, increasing the size of the depletion layer and the diode's resistance. High resistance stops the current from flowing through the diode, turning it into an insulator.</p> <p>That said, a large current can suddenly pass through the diode when a certain voltage level is applied. This level is known as the <a target="_blank" href="https://www.sciencedirect.com/topics/chemistry/breakdown-voltage" rel="noopener">breakdown voltage</a>. Above the breakdown region, even a small change in the voltage will cause a large change in the current.</p> <p>The forward bias and reverse bias properties of a semiconductor diode cause it to have an electrical conductivity value that's between the values of a conductor (very high) and insulator (zero or close to zero). An example of a conductor would be a metal like iron, whereas plastic, glass and rubber are types of electrical insulators.</p> </section> <section class="section main-article-chapter" data-menu-title="Types of diodes"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Types of diodes</h2> <p>Many types of diodes are available depending on their construction and application.</p> <h3>Rectifier</h3> <p>One of the most common types of diodes is the <a href="https://dynapower.com/rectifiers-nearly-everything-you-need-to-know/" target="_blank" rel="noopener">rectifier</a>. A rectifier is a P-N junction diode constructed of suitable semiconductor materials for the P-region and N-region. It lets the current flow only in the forward direction, i.e., when it is forward biased.</p> <figure class="main-article-image half-col" data-img-fullsize="https://www.techtarget.com/rms/onlineimages/photo_of_different_kinds_of_diodes-h.png"> <img data-src="https://www.techtarget.com/rms/onlineimages/photo_of_different_kinds_of_diodes-h_half_column_mobile.png" class="lazy" data-srcset="https://www.techtarget.com/rms/onlineimages/photo_of_different_kinds_of_diodes-h_half_column_mobile.png 960w,https://www.techtarget.com/rms/onlineimages/photo_of_different_kinds_of_diodes-h.png 1280w" alt="Image of various types of diodes." height="174" width="279"> <figcaption> <i class="icon pictures" data-icon="z"></i>There are many types of diodes -- specialized electronic components -- each with their own characteristics. </figcaption> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <h3>Zener diode</h3> <p>A Zener diode operates in reverse bias, letting the current flow in reverse, resulting in the Zener effect. The diode is also known as a breakdown diode since its cathode conducts current in reverse when a specific breakdown or threshold voltage is applied.</p> <p>Until the voltage is applied, the Zener diode will not conduct electricity from the cathode, connected to a positive voltage, to the anode, usually connected to the ground. The property of Zener diodes to enable reverse current flow provides a stable reference voltage. This quality makes Zener diodes suitable for use in power supplies, <a href="https://www.techtarget.com/whatis/definition/surge-suppressor-surge-protector">surge suppressors</a>, voltage shifters, waveform clippers and voltage regulators.</p> <h3>Schottky diode</h3> <p>A Schottky diode, also known as a hot carrier diode or low voltage diode, is formed by joining a semiconductor with a metal like platinum or aluminum. It has a lower forward voltage drop -- which occurs at low currents -- than conventional P-N junction diodes, and there are no stored charges in the junction, resulting in fast switching action. This makes Schottky diodes ideal for rectification applications, and for use in diode ring mixers and solar cell applications.</p> <p>A key drawback of Schottky diodes is that they have a large reverse saturation current, so they cannot withstand high voltage. Also, they have a narrow depletion zone, causing high leakage currents.</p> <h3>Laser diode</h3> <p>A laser diode converts electricity into light that's emitted in a tight beam. The conversion happens when an electrical current is applied to the diode, causing the electrons and holes in a P-N junction to recombine and release energy. This energy is in the form of photons of coherent light of a specific phase and wavelength. Laser diodes have numerous applications, including optical drives, barcode readers, laser printing and scanning, telecommunications and <a href="https://www.techtarget.com/searchnetworking/definition/fiber-optics-optical-fiber">fiber optics</a>, engraving, and laser pointers.</p> <h3>Light-emitting diode</h3> <p>Like laser diodes, LEDs generate light on the application of an electrical current. However, there are several differences between LEDs and laser diodes.</p> <p>Light-emitting diodes provide incoherent light that spreads out, whereas the light emitted by laser diodes is coherent and travels in a dense, straight beam. LEDs also have lower power output, are less expensive, and have larger intrinsic and active regions than laser diodes. Common applications of LEDs include lighting, displays and remote controls.</p> <p><em>Learn more about the <a href="https://www.techtarget.com/searchcio/feature/CHIPS-Act-moves-the-needle-on-US-chip-manufacturing">state of the U.S. semiconductor market</a> and read about <a href="https://www.techtarget.com/searchunifiedcommunications/answer/Do-you-really-need-4K-video-resolution-in-the-enterprise">4K video resolution in the enterprise</a>.</em></p> </section> A diode is a specialized electronic component, specifically a two-terminal semiconductor device, with two electrodes called the anode and the cathode. https://cdn.ttgtmedia.com/visuals/digdeeper/5.jpg https://www.techtarget.com/whatis/definition/diode Tue, 03 Dec 2024 12:00:00 GMT <p>MVS (Multiple Virtual Storage) is an <a href="https://www.techtarget.com/whatis/definition/operating-system-OS">OS</a> from IBM that continues to run on many of IBM's <a href="https://www.techtarget.com/searchdatacenter/definition/mainframe">mainframe</a> and large <a href="https://www.techtarget.com/whatis/definition/server">servers</a>. MVS is said to be the OS that keeps the world going. The same could be said of its successor systems, OS/390 and <a href="https://www.techtarget.com/searchdatacenter/definition/z-OS">z/OS</a>.</p> <p>The payroll, accounts receivable, transaction processing, database management and other programs critical to the world's largest businesses are usually run on an MVS or successor system. Although MVS has often been seen as a monolithic, centrally-controlled information system, IBM has repositioned it (and successor systems) as a "large server" in a network-oriented distributed environment, using a <a href="https://www.techtarget.com/searchsoftwarequality/definition/3-tier-application">3-tier application</a> model.</p> <section class="section main-article-chapter" data-menu-title="The evolution of MVS to OS/390 and z/OS"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The evolution of MVS to OS/390 and z/OS</h2> <p>The follow-on version of MVS, OS/390, no longer included the "MVS" in its name. However, the term "MVS" will continue to be used for some time because MVS represents a certain epoch and culture in the history of computing and many older MVS systems are still in operation.</p> <p>Since OS/390 also comes with <a href="https://www.techtarget.com/searchdatacenter/definition/Unix">UNIX</a> user and programming interfaces built in, it can be used as both an MVS system and a UNIX system at the same time. A more recent evolution of MVS is z/OS, an OS for IBM's zSeries mainframes. MVS systems run older applications developed using <a href="https://www.techtarget.com/searchitoperations/definition/COBOL-Common-Business-Oriented-Language">COBOL</a> and, for transaction programs, <a href="https://www.techtarget.com/searchdatacenter/definition/CICS">CICS</a>. Older application programs written in PL/I and <a href="https://www.techtarget.com/whatis/definition/FORTRAN-FORmula-TRANslation">FORTRAN</a> are still running. Older applications use the <a href="https://www.techtarget.com/searchdatacenter/definition/VSAM">Virtual Storage Access Method</a> access method for file management and Virtual Telecommunications Access Method for telecommunication with users. The most common program environment today uses the <a href="https://www.techtarget.com/searchwindowsserver/definition/C">C</a> and <a href="https://www.techtarget.com/searchdatamanagement/definition/C">C++</a> languages. <a href="https://www.techtarget.com/searchdatacenter/definition/DB2">DB2</a> is IBM's primary <a href="https://www.techtarget.com/searchdatamanagement/definition/RDBMS-relational-database-management-system">relational database management system</a>. <a href="https://www.theserverside.com/definition/Java">Java</a> applications can be developed and run under OS/390's UNIX environment.</p> <p>MVS is a generic name for specific products that included MVS/SP (MVS/System Product), MVS/XA (MVS/Extended Architecture) and MVS/ESA (MVS/Enterprise Systems Architecture). Historically, MVS evolved from OS/360, the OS for the System/360, which was released in 1964. It later became the OS/370 and the System/370. OS/370 evolved into the OS/VS, OS/MFT, OS/MVT, OS/MVS, MVS/SP, MVS/XA, MVS/ESA, and finally OS/390 and then z/OS. Throughout this evolution, application programs written for any OS have always been able to run in any of the later OSes -- also known as forward compatibility.</p> <p>An MVS system has a set of basic and optional products. Customers can choose the set of functions they need. The main user interface in MVS systems is Time Sharing Option. The Interactive System Productivity Facility is a set of menus for compiling and managing programs and for configuring the system. The main work management system is either Job Entry Subsystem 2 or 3 (JES2 or JES3). Storage management is performed by the Distributed File Storage Management Subsystem. MVS is more complex and requires much more education and experience to operate than smaller server and personal computer OSes.</p> <p>The <i>Virtual Storage</i> in MVS refers to the use of <a href="https://www.techtarget.com/searchstorage/definition/virtual-memory">virtual memory</a> in the OS. Virtual storage or memory enables a program to have access to the maximum amount of <a href="https://www.techtarget.com/searchstorage/definition/memory-card">memory</a> in a system even though this memory is shared among more than one application program. The OS translates the program's <a href="https://www.techtarget.com/whatis/definition/virtual-address"><i>virtual</i> address</a> into the real physical memory address where the data is located. The <i>Multiple</i> in MVS indicates that a separate virtual memory is maintained for each of multiple task partitions.</p> <p>Other IBM OSes for their larger computers include or have included: the Transaction Processing Facility -- used in some major airline reservation systems -- and <a href="https://www.techtarget.com/searchitoperations/definition/host-virtual-machine-host-VM">VM</a>, an OS designed to serve many interactive users at the same time.</p> </section> MVS (Multiple Virtual Storage) is an OS from IBM that continues to run on many of IBM's mainframe and large servers. https://cdn.ttgtmedia.com/visuals/digdeeper/3.jpg https://www.techtarget.com/searchdatacenter/definition/MVS Thu, 10 Oct 2024 00:00:00 GMT 60 webmaster@techtarget.com