Why would I consider Open Compute server platforms in my enterprise? Isn't this just for the biggest data centers?
Every data center operator understands the value of standardization. Picking one hardware platform, software platform, or even a business policy makes everyday business much easier to run. Unfortunately, many data centers face challenges in integrating and maintaining platforms that are constantly evolving and introducing proprietary, vendor-driven features that may not integrate with other systems in the environment. Or, it may lead to vendor lock-in, which does not serve business interests, wastes money and time, and makes the environment costlier to service.
Several years ago, Facebook promoted the Open Compute Project (OCP) to develop a series of technology initiatives intended to create systems with better power efficiency and reliability and that are easier to use and maintain. The goal was to provide computing power without all the proprietary bells and whistles that hamper integration and product choices.
This is generally accomplished through the open release of project documentation, allowing any system manufacturer to build systems that adhere to Open Compute's functional and mechanical standards. For example, Open Compute has established custom and high availability motherboards using AMD, Intel, and ARM processors.
This was first published in April 2013