Intel Corp. has been chosen as a yet another company to create an exascale supercomputer prototype for the Defense Advanced Research Projects Agency (DARPA). The company will cooperate with SGI and will compete against Nvidia Corp., who had also been chosen to explore exascale supercomputing technology for DARPA.
Building an exaFLOPS machine is a huge challenge. Even Advanced Micro Devices and Intel Corp. – whose x86 central processing units (CPUs) power the majority of supercomputers – admit that construction of a machine capable of performing quintillion floating point operations per second (1018 exaFLOPS) or more using x86 chips is hardly an executable task. The ability to achieve projected performance gains is limited by significant power consumption, architectural and programming complexity issues. To exploit available technological advances fully, highly programmable high performance computers must be developed that require dramatically less energy per computation. The goal of DARPA’s ubiquitous high performance computing (UHPC) program is to re-invent computing and develop radically new computer architectures and programming models that deliver 100 to 1000 times more performance, and that are easier to program than current systems.
Prototype UHPC systems are expected to be complete by 2018. The four performers selected to develop UHPC prototype systems are Intel, Nvidia, Massachusetts Institute of Technology's computer science and artificial intelligence laboratory, Boston and Sandia National Laboratory. Georgia Institute of Technology, was selected to lead an applications, benchmarks and metrics team for evaluating the UHPC systems under development.
"SGI is pleased to partner with DARPA and Intel on this groundbreaking program that promises to have far reaching implications for the supercomputing industry. To achieve the objectives of the DARPA UHPC program, we need to rethink how we are designing and building supercomputers. We look forward to a collaborative research environment that will result in major advances that redefine our industry.," said Dr. Eng Lim Goh, senior vice president and chief technology officer at SGI.
Interestingly, but Nvidia will also work with SGI to construct an exascale computer prototype. As a result, SGI will build two prototypes based on competing architectures: Nvidia's proprietary highly-parallel graphics processors as well as Intel's MIC architecture. One of the intrigues of the UHPC is what hardware architectures will be preferred by the MIT and Boston & Sandia National Laboratory since those institutions do not develop commercial hardware and are likely to prefer the most perspective components.
Tags: Intel, SGI, DARPA, x86, GPGPU, Nvidia, MIC
Comments currently: 1
Discussion started: 01/22/11 06:05:22 AM
Latest comment: 01/22/11 06:05:22 AM
SGI or NVidia. Realistically, the winner will be the one who can produce the fastest chip that conforms with the most efficient form factor that fits the ideal Power, Cooling and Network solution.
Intel, of course, are is only choice!
For more understanding:
I created the e-mail address when I named my design in November 1999. So if anyone has a problem with patents, talk to hotmail.com, because I can't remember when I thought to use it as my e-mail address.
Also, why do people keep hearing only the big numbers and not the little ones that matter. Relating to network latency, node to node synchronization and of course, cost! I may not have known very much about how a Beowulf worked, but I did recogise that a tolerance measured in femtoseconds could be taken advantage of. Which is only possible, due to the cost effective nature of free space optics.
Everyone who has the power to make real change, talk about 'Radical ideas', but only if it doesn't require complete re-tooling or work force redundancy.
Which are decisions I would not wish on anyone!
01/22/11 06:05:22 AM]
Add your Comment
Enter your username and e-mail address. Password will be sent to you.