Earlier this week, we reported that the Virginia Tech supercomputing cluster project built from 1,100 Power Mac G5s, was pushing some 7.41 million floating point operations, or TFlops. This number fell far short of the theoretical maximum for the system of 17.6 TFlops that was reported two weeks ago, but is still enough to possibly land it in the Top 5 fastest systems. Today, an article in the Times of India is reporting that the team at Virginia Tech has boosted performance for Big Mac, as it is called, to 8.164 TFlops, a rating that could bump it up to3rd place in the Top500.orgis listing of the worldis fastest supercomputers.
That said, the article mentions yet another new system that may also crest its way into the Top 5 supercomputers at 11.8 TFlops, which would likely push Big Mac back down to #4. The numbers that are being tossed around currently are not "official" numbers, and many operators of these supercomputers and supercomputing clusters are busily tweaking their systems for the next official ranking, which should be announced next month. In the case of Big Mac, Virginia Techis team is still working on increasing the efficiency of the cluster. From the Times of India:
In benchmark tests last week, the Virginia Tech supercomputer, powered by 2,200 IBM microprocessors, could compute at 7.41 trillion floating point operations (teraflops) a second. That would place it fourth in the list of fastest supercomputers in the world, according to The New York Times, which first reported the story.
But by Wednesday evening, when Varadarajan spoke to this correspondent, the machine had been powered up to 8.164 teraflops, pushing it up to a possible third place. "And itis only now operating at perhaps 50 per cent efficiency," Varadarajan said in the interview.
The main point of the full article was to profile the India-born engineer in charge of the Virginia Tech program, but it also contains more general information on Big Mac. We recommend it as an interesting read.