The U.S. government plans to install 19,000 Tesla K20 modules from Nvidia into its Titan supercomputer at Oak Ridge National Laboratory in Tennessee. Once the machine is upgraded, it will be eight times quicker than it is now, carrying out an estimated 25 quadrillion floating point operations per second (25 petaflops).
GPUs are highly efficient at carrying out parallel processing tasks, where a process can be broken down into many parts and computed simultaneously since the outcome of a single computation does not determine the input of another.
Supercomputers are switching to this hybrid computing system, splitting work up between thousands of CPUs, and thousands of GPUs, depending on which will hand the task in a more efficient manner. The Titan supercomputer is used to help develop more efficient energy systems (such as vehicle engines), to model changes in climate and other complex tasks.
"If you take a look at scientific applications, 99% of the operations can be done in a highly parallel manner, and that can be done much more efficiently by large numbers of very simple GPU processors than on a traditional CPU burning a lot of power trying to make a single thread go fast," Steve Scott, Nvidia's chief technology officer, told the BBC.
"I liken CPUs to a Tour de France where a whole team of trucks and support staff are built around one athlete to help them win the race - a lot of energy making one thing go fast - as opposed to a parallel throughput approach where you make thousands of things in aggregate go fast."
When completed, the Titan will be more than twice as powerful as the K Computer, which currently occupies the top spot on the TOP500 list of supercomputers.
Written by: James Delahunty @ 19 May 2012 20:58
