Made in China – World’s fastest supercomputer

Thursday, 28 October, 2010

China has used Intel and Nvidia technologies to build the world’s fastest supercomputer Tianhe-1A, at the National University of Defence Technology in Tianjin.

Defense university builds China’s fastest supercomputer, Xinhuanet

“The title has gone to China’s Tianhe-1A supercomputer that is capable of carrying out more than 2.5 thousand trillion calculations a second.

To reach such high speeds the machine draws on more than 7,000 graphics processors and 14,000 Intel chips.

[...] Tianhe-1A is unusual in that it unites thousands of Intel processors with thousands of graphics cards made by Nvidia.

The chips inside graphics cards are typically made up of small arithmetical units that can carry out simple sums very quickly. By contrast, Intel chips are typically used to carry out more complicated mathematical operations.

The machine houses its processors in more than 100 fridge-sized cabinets and together these weigh more than 155 tonnes.”

China claims supercomputer crown, BBC News

China builds world’s fastest supercomputer, UK Telegraph

- Is China a supercomputer threat? (Q&A), CNet


P != NP (an update)

Monday, 23 August, 2010

I very much agree with this quote in “A Tale of A Serious Attempt At P≠NP” by Richard J. Lipton – (August 15, 2010)

“… I learned that the Internet may not replace referees yet, but it can raise many important questions in a positive and constructive manner.”

With respect to the P != NP “proof”, I am not optimistic. Quoting “The problem of P versus NP” (full article enclosed below),

“Mr. Deolalikar has submitted a revised proof, but Dr. Cook believes the jig is up. Does he believe P vs. NP will ever be solved? “Yes, I do, I do – eventually. But not very soon. I think it really is a hard problem. It’s becoming increasingly clear, because so many top mathematicians have tried to solve it.””

The problem of P versus NP – Kate Allen From Saturday’s Globe and Mail – Published on Friday, Aug. 20, 2010 10:28PM EDT

In 1971, Dr. Stephen Cook, a young University of Toronto professor in the fledgling field of computer science, posed a theoretical problem so intractable it has become the subject of a $1-million prize. Since then, only a handful of credible solutions have been posed. All of them fell short. This month, one man caused a commotion after he announced to experts in the field that he had solved the problem known as P vs. NP.

The proof

Vinay Deolalikar, a Delhi-born mathematician at Hewlett-Packard, sent Dr. Cook and two dozen other experts in the field an e-mail, writing, “I am pleased to announce a proof that P is not equal to NP, which is attached.” The paper was more than 100 pages long. Dr. Cook was excited. In 40 years, “I can think of only a couple of other attempts of people who’ve thought they’ve proved it,” Dr. Cook said. “Most of them you can dismiss very easily – they’re not really mathematicians. But this one was much more serious.”

The problem

Clay Mathematics Institute, the Cambridge, Mass.-based academy that offers the $1-million prize, describes the problem with this example: Imagine you are trying to figure out housing for 400 university students. Read the rest of this entry »


Recent P != NP excitement & activities

Friday, 13 August, 2010

P = NP or P != NP is an important and fundamental problem in mathematics and computer science. Vinay Deolalikar, a research Scientist at HP Labs, has circulated a paper that he thinks prove P != NP (see an older version here). Some researchers in the field have pointed out some general issues (fatal?) and created a really impressive wiki about Vinay’s paper to track and document the various discussions for the public to try to follow. Cool stuff.

Love this quote,

We think that Vinay Deolalikar should be thanked for thinking about this difficult problem, and for sharing his ideas with the community. Whether he got it right or not, he has tried to add to our understanding of this great problem. We need more people working on hard problems. If no one does, then they never will be solved.

On a personal note, I love to see P =? NP resolved in my life time since my first exposure to the problem in 1990 in Prof. Stephen Cook‘s undergrad computation complexity class at U of T. My sense this morning is that Vinay may not have cracked the problem yet, I do hope he or other researchers (individually or as a team) will finally crack it. Good luck to them all.

[HT Schneier Gomes]

P.S. Incidentally, Michael Nielsen (the person hosting the above wiki) is working on a book, to be published by Princeton University Press in 2011, that describes both the great potential of online collaboration tools “to improve how science is done, and the challenges that must be overcome to fully realize that potential”. A preview of some themes of his book may be found in his essay about The Future of Science.

I also noticed this interesting wiki for “polymath projects – massively collaborative online mathematical projects“.


Follow

Get every new post delivered to your Inbox.

Join 678 other followers

%d bloggers like this: