News Stay informed about the latest enterprise technology news and product updates.

Linux cluster no crash test dummy for Volvo

News that Volvo is implementing a Linux high-performance computing cluster bodes well for automotive safety. But the bigger story is that IBM has taken the technology from university environments to commercial applications.

Computer-based automotive crash simulations haven't made significant headlines since the 1990s, when vendors like Sun Microsystems Inc. and Hewlett-Packard Co. were providing workstation-based simulation software to automotive players like Audi and BMW.

With that information in hand, many authorities in the automotive circuit may have been scratching their heads this week about the relevance of an IBM announcement that had the Volvo Car Corp. purchasing a high-performance computing (HPC) platform to conduct automobile crash simulations.

However, Pund-IT Research principal analyst Charles King said the recent deal between the Swedish auto maker and IBM is one worth noting, especially if you frequent the Linux circuit.

The HPC platform IBM is scheduled to deliver arrives in the form of 150 IBM eServer 325s with AMD Opteron processors, which will be combined with Volvo's already existing eServer xSeries 335 and pSeries 655 HPC platform.


Linux clusters for HPC -- Why and why not?


Cray CTO -- Supercomputers outshine Linux clusters, part 1

IBM said it expects the combination to become one of the industry's fastest Linux clusters based on statistics.

King said he believed the purchase was representative of several important factors in the industry.

First off, he said, IBM has piled additional evidence that the Linux operating system is now the OS environment of choice in the "high-stakes world or automobile manufacturing."

"Until recently, auto manufacturers largely relied on Unix-based solutions from vendors, including SGI, Sun and HP," King said. "Along with its Volvo and GM supercomputer wins, IBM has also built Linux-based clusters for Ferrari and Daimler Chrysler for design/analysis processes."

On a side note, King said the Volvo deal also bodes well for AMD, who had enjoyed notable success on the Opteron platform in HPC, but will seemingly enjoy more success -- due to public demand for 64-bit x86 processors -- in the volume server space.

"The IBM deal offers [AMD] an opportunity to refocus on one of Opteron's core value-adds," King said.

Finally, King explained, the deal provides another example of the inroads supercomputing products have made away from their historical niche in university and research lab environments and into commercial applications.

"Overall, we see the affordability and flexibility of clustered supercomputing systems as prime drivers of their success, but how vendors position such solutions for vertical markets is critical," he said.

King said IBM, in particular, focuses a great deal of attention and energy on defining the value of supercomputing for specific-industry applications, including automotive design and manufacturing, oil and gas exploration, pharmaceutical development and film production.

By offering cost-effective commercial solutions, such as Volvo's new Opteron system, King said, "IBM demonstrates once again that it can apply the power of supercomputing solutions to real-world problems."

Dig Deeper on Linux servers

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.