Become a fan of Slashdot on Facebook


Forgot your password?
Supercomputing Technology

China Makes World's Fastest Supercomputer 222

shmG writes "China has replaced the United States as the maker of the world's fastest supercomputer. A Chinese research center has made the world's faster super computer — named as Tianhe-1A, which was released at a national conference on high-performance computers (HPC) in China. Made at a cost of over $88 million, Tianhe-1A is theoretically able to do more than 1 quadrillion calculations per second (one petaflop) at peak speed. Tianhe-1A 's peak performance reaches 1.206 petaflops, and it runs at 563.1 teraflops (1,000 teraflops is equal to one petaflop) on the Linpack benchmark."
This discussion has been archived. No new comments can be posted.

China Makes World's Fastest Supercomputer

Comments Filter:
  • by fluffy99 ( 870997 ) on Thursday October 28, 2010 @08:41AM (#34048108)

    But does it run (Red Flag) Linux?

  • Maybe this is the bitch-slapping the US needs to pull it's head out of it's ass, and start doing the things it needs to do to be seriously competitive again.

    Meh. Who am I kidding?

    • by hsmith ( 818216 ) on Thursday October 28, 2010 @08:47AM (#34048174)
      Yeah, lets borrow more money from the Chinese Government so we can build a useless supercomputer to outdo them - just to say we did it! Thanks, grandkids!
      • Where did I use the word "government"?

        • Re: (Score:3, Insightful)

          by jgagnon ( 1663075 )

          Hint: cynics don't require facts to be cynical.

        • by hsmith ( 818216 )
          I don't really see what company would make the investment - for what? There are plenty of supercomputers in the US already. If they needed to perform some processing, they would leverage those existing investments for a lot less than what building their own would cost - or they would use the chinese one because it is the "fastest" for the time being. There is little incentive to "build one because the chinese have a faster one" in the commerical market.
          • It wasn't the US gov that built the largest sky scrapers in the world, yet they were all built in the US until recently... who built them? Chrysler... Sears... etc.

            Companies most certainly do need massive supercomputers. Oil companies are a good target, and Exxon happens to be the largest corp in the world.

            As to using surplus supercomputers being cheaper... well supercomputers tend to be rather task specific. Additionally, since you're completely generalizing, i'd have to say you're making that up on

      • by dbet ( 1607261 )
        We're already using the money we borrow from China to protect Asia from China.
      • Re: (Score:2, Interesting)

        There is nothing useless about a supercomputer. Oak Ridge National Lab has over a billion dollar budget [] each year and huge portions of that budget relies on the availability of high performance computing resources. (Not to mention all the other national labs []) HPC supports research in areas like energy conservation, new power sources, bioinformatics, material science, weapons simulations, engineering, and computer science. Applications range from freeing ourselves of fossil fuel reliance to designing mat
        • "weapons simulations"

          Useless, that is.

        • by mcgrew ( 92797 ) *

          HPC is the reason our grandkids will have a longer average lifespan.

          Then why isn't ours the longest in the world? Yes, supercomputing is very important for all sorts of scientific research, including medicine, but if you want a longer average lifespan, simply make workplaces less dangerous and give all of us good health care, like the nations with longer lifespans than ours have.

      • I saw an ad that says - in a menacing voice - "and now they for for us - and I thought, Isn't that exactly what is supposed to happen? The Chinese have been collecting dollars and not contributing to demand - exactly what is needed is for China to spend and put US companies to work, Well, here it is: US working for them, Intel and nvidia making some coin, paying some employees who pay some taxes, in a virtuous circle.

      • by mlts ( 1038732 ) *

        Since when did supercomputers become useless? They might not be as mainstream as bygone days when supercomputers in the past were the only things that could render stuff well for cool pictures, but they have a critical need in a lot of research, especially complicated models with a ton of variables to consider. Most variables are floating point calculations so the average integer processing unit wouldn't help much.

        Of course, a supercomputer won't be that great at regular integer stuff (if you want your BS

      • by mcgrew ( 92797 ) *

        Useless supercomputer?

        I think you're on the wrong site...

    • by WrongSizeGlass ( 838941 ) on Thursday October 28, 2010 @08:52AM (#34048218)
      Sadly, China needed to build this computer simply to calculate the interest on the US debt in realtime.
    • Re: (Score:3, Insightful)

      by aliquis ( 678370 )

      What are you suggesting?

      Growing population by a factor of 5?

      Decrease salaries?

      Spend lots of the money on small high image projects while most of the rest of the country remain poor?

      I do understand that they will eventually catch up, but in the mean time you Americans are way ahead of the average Chinese.

    • by betterunixthanunix ( 980855 ) on Thursday October 28, 2010 @09:14AM (#34048498)
      Competitive in what, exactly? We have many supercomputers in the USA; we have no idea what to do with them, though, and many of them are spending a lot of time idle. Some supercomputers are now being rented out to investors, because the people the computers were built for -- the scientists -- are not using enough computer time.

      What we really need to do is look at the state of research in this country. Also, maybe if we had a more solid economic base, one in which we solve the trade imbalance by exporting real goods rather than copyrights, we could spend more money on science and supercomputing. Oh well, in your words, "who am I kidding?"
    • This is something that used to bug me, but then I realized something important. I just do like a large amount of other people from unstable regions have done for the last 30 years, and flee like a rat from a sinking ship. The US fails and I just get a tech job in a different (and much more stable) country. I work for a multinational corporation: Transfers are nontrivial, but relatively easy. It's more a question of motivation.
    • The 3 gorges will have nothing on us. That'll show 'em!

      And if destroying the natural beauty isn't enough, I'm sure we can find some priceless pueblo or ancient burial mound, and turn it into a cooling pond for a giant nuclear plant. Take that, China!

  • by digitaldc ( 879047 ) * on Thursday October 28, 2010 @08:47AM (#34048172)
    And the fastest social and economic downturn is in America...coincidence?
    • And the fastest social and economic downturn is in America...coincidence?

      Yeah, but this is just their firewall. Wait until you see the systems behind it.

      As for the China's climb to the top, everyone hated U.S. policies were the big dog. Well, I hope those folks enjoy the ride down, cause it will not be pretty. I'd rather be on the top of the hill and hated, then be nice and shit all over.

    • China is developing country where as the US is a developed one.

      It's much easier to invest and build infrastructure where none exists in the first place (clean slate) than it is to uproot your existing stuff and replace it with newer. For the later, it's also much more difficult to justify the future break-even point of such an investment. People want results NOW over and beyond a simple TCO calculation.

    • And the fastest social and economic downturn is in America...coincidence?

      You're right! Clearly America needs a heavy dose of green tea and innards, stat!

    • by the_humeister ( 922869 ) on Thursday October 28, 2010 @10:23AM (#34049488)

      Oh my. When have I heard this before? Oh yes, back in the 1980s when there was panic and hyperbole over Japan, Inc. overtaking the USA in everything. How did that pan out exactly? I don't see how the current situation with China is any different.

      • Well for starters, look at the size of China vs. Japan...

      • by mlts ( 1038732 ) * on Thursday October 28, 2010 @02:06PM (#34053320)

        China is way different from Japan. For starters, it has resources, and it can play the game any way it wants to. Japan could only play hardball economically. China can at any time choose to overrun Korea and Taiwan at any time if they choose to, and the only recourse would either be a hard fought conventional war, or a nuclear exchange.

        China can fight dirty. Japan cannot. And China is good at fighting dirty, because they "won" two wars (Korea and Vietnam) by proxy, sending in men and materials to do what the native population couldn't. If China chose to, they could easily turn up the heat in other areas hostile to the US by sending in troops and munitions. China could hand Iran the tools to seize control of the Strait of Hormuz and there would be nothing the US could do about it except engage in another theater of war which would be unwinnable.

  • Finally they will be able to computerize the national census procedures.

  • I mean, who ever expected that Skynet would speak Chinese?
  • After observing Red Flag, Loongson, and the basic nature of Chinese hardware, I predict we're going to shortly see an "oh wait, they were lying, it's 200 teraflops of American hardware" come down the pipe.

    I wouldn't mind being wrong, though.

  • Fastest?! (Score:5, Informative)

    by aaarrrgggh ( 9205 ) on Thursday October 28, 2010 @09:05AM (#34048388)

    Oak Ridge (Jaguar):
    Cores Rmax(GFlops) Rpeak(GFlops) Nmax Nhalf
    224162 1759000 2331000 5474272 0

    Seems faster by a good margin.

    • Re: (Score:3, Informative)

      The Chinese computer has the fastest theoretical "peak" performance, but how well that translates to actual operational efficiency really depends on how well they are actually able to utilize the GPUs. This computer makes massive use of the GPUs which sort of gives it an architecture similar to the earth simulator, ie a massive # of vector processors supplemented by some scalar cpus. GPUs have a lot more memory bandwidth restrictions when compared to the general purpose vector CPUs used in the earth simul
  • Numbers Correction (Score:4, Informative)

    by airwick ( 121385 ) on Thursday October 28, 2010 @09:07AM (#34048400) Homepage

    The slashdot summary has the wrong numbers. The actual article which slashdot quotes is contradictory. Its starts by saying:
    "Tianhe-1A has set a new performance record of 2.507 petaflops, as measured by the Linpack benchmark, making it the fastest system in China and in the world today."
    and then one paragraph later it gives the same numbers as the slashdot summary.

    Other articles (from other sites) are claiming theoretical peak performance of 4 Petaflops (from an Nvidia source) and sustained petaflops of 2.5.

    • Re: (Score:2, Informative)

      by fahlesr1 ( 1910982 )

      Blue Waters [] is supposed to have a theorectical peak performance of 10 petaflops and sustained performance of 1. Personally I doubt that Tianhe-1A can sustain 2.5, I think the half petaflop number is probably more accurate.

      Its easy to throw lots of CPUs together, its much harder to keep them all busy.

  • From TFA:
    "... Tianhe-1A has set a new performance record of 2.507 pataflops, as mesured by the Linpack benchmark ... Tianhe-1A is theorectically able to do more than 1 quadrillion calculations per second(one petaflop) at peak speed. Tianhe-1A's peak performance reaches 1.206 petaflops...

    So does it do 1, 1.206 or 2.507 petaflops ?

  • China lies. (Score:5, Insightful)

    by rafter109 ( 1319819 ) on Thursday October 28, 2010 @09:19AM (#34048544)
    There is absolutely no consistency in numbers in this story. Some measurements show this computer to be about 45% slower than the Cray XT5 and some show it to be faster. Given China's history of arbitrarily throwing out numbers to try to prop themselves up in the international community I cannot accept this as fact without some sort of independent verification. If China has in fact created the worlds fastest supercomputer, I congratulate them on a job well done. But I am still skeptical about this story. Sounds like my government (US) is just looking for an excuse to spend billions more on a new supercomputer.
  • Fire this reporter (Score:5, Informative)

    by Orgasmatron ( 8103 ) on Thursday October 28, 2010 @09:26AM (#34048622)
    The article starts by claiming 2.507 petaflops, but gives no mention if that is Rmax or Rpeak. We have to assume that it is Rmax, since 2.5 petaflops is no big deal in terms of Rpeak.

    Unfortunately, then the article lists both Rpeak and Rmax. But the numbers quoted seem to be for Tianhe-I (#7 on the top 500 list), not Tianhe-IA (not currently listed). Wikipedia table of the top 10 []

    Oh, and it gets better. The article claims that Tianhe-IA has 7,168 GPUs and 14,336 CPUs. Very strange, since the Tianhe-I has 71,680 CPU/GPU pairs.

    My guess is that China doubled up their Tianhe-I computer and swapped out for newer GPUs, then named the new thing Tianhe-IA (this is pretty normal when competing for top500 spots). I'm going to go with 143,360 Xeon/M2050 pairs. Either that, or the Chinese found a way to overclock 10% of their chips into the 20+ GHz range and threw out the rest.
  • by sarkeizen ( 106737 ) on Thursday October 28, 2010 @09:30AM (#34048682) Journal
    Supercomputing is largely a solved a regression analysis on the variables on the Top 500 sometime. Primarily it's a function of the number of cores. I've told people in my own workplace - you want a machine on the Top 500 then write me a cheque - making a supercomputer isn't the feat of skill or engineering that it was in the days of Cray. This doesn't even touch on what the hell you use these things for the problem space for parallel processing is clearly smaller than that of serial processing. Add to that the assertion that the number of useful applications drops off steeply as you add more cores (at some point you are left with only the "embarrassingly parallel" ones) and creating the largest supercomputer in the world is akin to saying you are creating the least useful computer in the world. Not to mention probably one of the least power efficient, highest maintenance costs, etc..
    • by JerkBoB ( 7130 )

      I love how using a few fancy words and writing assertively is the only thing necessary for getting a few "Insightful" mods.

      You don't know what the hell you're talking about. As someone who's been living and breathing HPC for years, I can say that supercomputing is most definitely NOT passe.

      It's true that supercomputers are only needed for a relatively small subset of computing problems, but that subset is pretty damned important to science. There are some problems which are just too huge to be handled wit

      • Call me back in a week when you've finished loading that 400TB dataset into S3. And after you've taken an eon to crunch it all, you'll need another week to dump the results back out.

        To be fair, if you're going to (have had to) wait for an eon for the processing, a week to (un-)load the data doesn't seem like that big of a deal.

      • Weird. The article was about building supercomputers - so was my comment. And the EC2 thing - man where did that come from? I'm one of the biggest critics of that technology. It is definitely NOT a replacement for large cluster environments even small highly parallel tasks the MD5 cracking - It's like 3x slower and 2x as expensive than using a high-end GPU.

        "It's OK to be ignorant about something outside of your field."

        How about being ignorant about a post you allegedly read?
  • What metric should one use to compare supercomputers? Using the amount of floating point operations performed per second (FLOPS) comes across as a little silly, because if you just pile together a sufficient amount of standard PC's, you should be able to top it (maybe Folding@home could be considered a bigger supercomputer). Therefore, I assume that the interconnectivity (and the speed related to it) of the CPU's should have something to do with it...

    • by Andy Dodd ( 701 )

      I'm fairly certain Linpack (the standard metric) requires decent connectivity if it is to scale.

  • Meh (Score:5, Funny)

    by csoto ( 220540 ) on Thursday October 28, 2010 @09:45AM (#34048882)

    The only reason it's so fast is because a half hour after you feed it data, it's hungry again...

  • his sources. As reported on CNN from Mashable []:

    Unveiled Wednesday at the Annual Meeting of National High Performance Computing (HPC China 2010) in Beijing, Tianhe-1A is the world's fastest supercomputer with a performance record of 2.507 petaflops, as measured by the LINPACK benchmark.

    Tianhe-1A was designed by the National University of Defense Technology (NUDT) in China, and it is already fully operational.

    To achieve the new performance record, Tianhe-1A uses 7,168 Nvidia Tesla M2050 GPUs and 14,336 Intel Xeon CPUs.

    It cost $88 million; its 103 cabinets weigh 155 tons, and the entire system consumes 4.04 megawatts of electricity.

    Tianhe-1A ousted the previous record holder, Cray XT5 Jaguar, which is used by the U.S. National Center for Computational Sciences at Oak Ridge National Laboratories.

    It is powered by 224,162 Opteron CPUs and achieves a performance record of 1.75 petaflops.

    According to Nvidia, Tianhe-1A will be operated as an open access system to use for large scale scientific computations.

    Just sayin...

  • Long ago I mis-understood what a "bit slice" processor really was... and the bit grid [] was born. If I'm right it's possible to build something to kick the current technology to the curb, running in a single rack.

    Due to the Von Neuman bottleneck, most of the transistors in a computer are in the RAM, which is idle except for the row/column being accessed at a given time. The bitgrid gets around this by building a grid of look up tables which operate on 4 bits in and 4 bits. It should be quite easy to build

    • by HuguesT ( 84078 )

      Look up "artificial retina", not the kind to restore vision to the blind, but the kind to perform computer vision tasks

      Very similar to what you are suggesting.

      • by ka9dgx ( 72702 )
        It's unfortunate that the article is behind a payway, as it does look quite interesting.
  • Call me when China invents something like Tang.

  • I'll bet it can't!

  • Tianhe-1A is theoretically able to do more than 1 quadrillion calculations per second (one petaflop) at peak speed. Tianhe-1A 's peak performance reaches 1.206 petaflops,

    But according to the news Jaguar is able to mosey along at 1.75 petaflops, and Tianhe is rolling along at 2.5. The petaflops barrier was broken a few years ago, so Tianhe is basically a product of Moore's law.

Adding manpower to a late software project makes it later. -- F. Brooks, "The Mythical Man-Month"