Slashdot Mirror


Japan Unveils Next-Generation, Pascal-Based AI Supercomputer (nextplatform.com)

The Tokyo Institute of Technology has announced plans to launch Japan's "fastest AI supercomputer" this summer. The supercomputer is called Tsubame 3.0 and will use Nvidia's latest Pascal-based Tesla P100 GPU accelerators to double its performance over its predecessor, the Tsubame 2.5. Slashdot reader kipperstem77 shares an excerpt from a report via The Next Platform: With all of those CPUs and GPUs, Tsubame 3.0 will have 12.15 petaflops of peak double precision performance, and is rated at 24.3 petaflops single precision and, importantly, is rated at 47.2 petaflops at the half precision that is important for neural networks employed in deep learning applications. When added to the existing Tsubame 2.5 machine and the experimental immersion-cooled Tsubame-KFC system, TiTech will have a total of 6,720 GPUs to bring to bear on workloads, adding up to a total of 64.3 aggregate petaflops at half precision. (This is interesting to us because that means Nvidia has worked with TiTech to get half precision working on Kepler GPUs, which did not formally support half precision.)

121 comments

  1. Pascal-based? by davidwr · · Score: 5, Insightful

    Am I the only one that thought "LISP machines, okay, but Pascal?"

    --
    Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
    1. Re:Pascal-based? by dougdonovan · · Score: 1

      ok, we're on the same page. japan, wake up.

    2. Re:Pascal-based? by cant_get_a_good_nick · · Score: 1

      I was thinking "Mr Wirth would be very happy..."

    3. Re:Pascal-based? by sycodon · · Score: 1

      Umm...is this the Pascal as in the programming language?

      --
      When Fascism comes to America, it will call itself Anti-Fascism, and tell you to give up your guns.
    4. Re:Pascal-based? by Crashmarik · · Score: 2

      Being a Pascal programmer I felt horribly bait and switched.

    5. Re:Pascal-based? by Anonymous Coward · · Score: 1

      The real question is, did anyone NOT think that?

      And, more importantly, just how calculated was that clickbait title, and who or what calculated it?

    6. Re:Pascal-based? by s.petry · · Score: 2

      Pascal has always been exceptional performing math. Makes sense for a super computer IMHO. The fact that I wrote pascal for my math degree is bonus!

      --

      -The wise argue that there are few absolutes, the fool argues that there are no probabilities.

    7. Re:Pascal-based? by Anonymous Coward · · Score: 1

      but Pascal [standardpascal.org]?

      It's more like Pascal or, alternatively, Pascal.

    8. Re: Pascal-based? by Anonymous Coward · · Score: 0

      At least they didn't try PROLOG again....

    9. Re:Pascal-based? by ShanghaiBill · · Score: 4, Informative

      Umm...is this the Pascal as in the programming language?

      Does anyone still program in Pascal? The last time I saw it on a resume was more than 20 years ago.

      Anyway, this supercomputer has nothing to do with the Pascal programming language. It is built using NVidia Pascal GPUs.

    10. Re:Pascal-based? by BitterOak · · Score: 1

      Pascal has always been exceptional performing math. Makes sense for a super computer IMHO. The fact that I wrote pascal for my math degree is bonus!

      Are you sure you're not thinking of Fortran? I've never heard that Pascal is unusually good at math compared to other general purpose programming languages.

      --
      If I can be modded down for being a troll, can I be modded up for being an orc, or a balrog?
    11. Re:Pascal-based? by Anonymous Coward · · Score: 0

      Yes, it's called Delphi.

      I love Pascal. In my opinion it is probably the best high level language ever created. It's so clean and structured. It forces good coding practices.

    12. Re:Pascal-based? by jimtheowl · · Score: 1

      Some people do. It is a very well defined language worth looking into even if only for academic reasons.

      Some would say that everything old is new again; good design merits consideration.

      I am not suprised by this. When faced with coming up with something better, why not choose something old (and proven) rather than try to create something new?

    13. Re:Pascal-based? by Lips · · Score: 1

      Inno Setup is an excellent free installer for Windows programs. It's scripting language is Pascal. http://www.jrsoftware.org/isinfo.php/

    14. Re:Pascal-based? by Anonymous Coward · · Score: 0

      Inno Setup is what gog.com uses for all of the games they sell.

    15. Re:Pascal-based? by sycodon · · Score: 1

      OK, thanks.

      I liked Pascal. It was very elegant.

      --
      When Fascism comes to America, it will call itself Anti-Fascism, and tell you to give up your guns.
    16. Re:Pascal-based? by s.petry · · Score: 1

      I programmed Fortran my first year, but after that we used Pascal. Pascal was not as fast for some operations, but in others was faster. C was no match, and yes I took a year of C and even a semester of Cobol just so I could say I did.

      --

      -The wise argue that there are few absolutes, the fool argues that there are no probabilities.

    17. Re:Pascal-based? by jbolden · · Score: 1, Redundant

      Well first off the super computers aren't about the Pascal language but the Pascal chip. I'd disagree that Pascal was all that proven out. It seemed very quickly to have had structural flaws which caused other languages to overtake it. Pascal was fairly low level yet it lacked good low level interfaces. Which is why it lost out to C. Pascal supports admit this and one of the main directions of Turbo Pascal / Delphi was to introduce into Pascal handling for lower level code (example partial compilation).

      If you think of Pascal as a higher level language where bad handling of low level code is acceptable it also wasn't competitive. Pascal is strongly typed but has a poor type system without abstractions. Making types difficult to work with under almost all conditions. It had poor handling of static vs. dynamic data including things like abstracting networks or file systems. There are lots of sacrifices in organization for ease in writing small compilers. A very good choice for early 1980s PC compilers that had to run off a floppy not a good choice since. The languages strictness on looping structures tended to result in duplicate code.

      Etc... Pascal was a partial success. But it died for good reasons.

    18. Re:Pascal-based? by squiggleslash · · Score: 1

      APK does. He's boasted several times his Hosts file manager is written in Delphi, which is apparently an Object Pascal implementation and the successor to Turbo Pascal.

      --
      You are not alone. This is not normal. None of this is normal.
    19. Re:Pascal-based? by angel'o'sphere · · Score: 1

      Sorry but eveything you write about Pascal is wrong. But I'm to lazy to correct it, you can read wikipedia.
      You clearly never used it, so why write that bullshit?

      Pascal, especially as UCSD P-System was for decades the most widely used 'OS' and programming language on the planet.

      --
      Cost free eBook I read (by iBook/Kobo/Amazon/ObookO/Gutenberg etc.): "The Green Odyssey" by Philip Jose Farmer.
    20. Re:Pascal-based? by angel'o'sphere · · Score: 1

      He meant Pascal, the human. Not Pascal, the language.
      And Pascal as languge is as good as C in doing math ... or basically every pro edural language.

      --
      Cost free eBook I read (by iBook/Kobo/Amazon/ObookO/Gutenberg etc.): "The Green Odyssey" by Philip Jose Farmer.
    21. Re:Pascal-based? by Anonymous Coward · · Score: 0

      Am I the only one that thought "LISP machines, okay, but Pascal?"

      And my second thought was:

      "Well, this is coming from the Tokyo Institute of Technology...

      Go TITs!"

    22. Re:Pascal-based? by jbolden · · Score: 0

      Huh? When? What year? The p-system is released around the same time as the microcomputer. Pascal never approaches the ROM/Basic/OS combo those systems defaulted to.

    23. Re:Pascal-based? by angel'o'sphere · · Score: 1

      You should read the relevant wikipedia articles :)

      --
      Cost free eBook I read (by iBook/Kobo/Amazon/ObookO/Gutenberg etc.): "The Green Odyssey" by Philip Jose Farmer.
    24. Re:Pascal-based? by jbolden · · Score: 0

      That's a longer answer than a single number or range of numbers.

    25. Re:Pascal-based? by jimtheowl · · Score: 1

      I don't know if other languages overtook Pascal because of 'structural flaws', but I myself adopted C over Pascal because at the time it was obviously better for raw power and speed. This is true for the processors and architectures that I have developed on (6502/x86/SPARC/AMD64).

      As much as I love C, it is not a good fit on other architectures such as stack machines. Burroughs machines (now UNISYS) use an extended version of ALGOL as a system language with impressive results, especially when it comes to security (not vulnerable to buffer overflows).

      I am a C programmer at heart, but have an argument with your assumption that Pascal (or related languages) is dead. I could be wrong, but I think you may be brushing off its merits based on the widespread architecture of today.

      Yet, you may have hit the nail when saying it is about the Pascal chip, but that doesn't exclude the language - it makes it a holistic part of the system design.

  2. Awww... by Anonymous Coward · · Score: 0

    And here I was hoping that the "Pascal" was the language and that maybe they were finally realizing that decades of language and OS development have been going in the completely wrong direction. What we need is a UCSD Pascal powered supercomputer.

  3. Yay! by sharkette66 · · Score: 0

    I can get out my Turbo Pascal box again!

    1. Re:Yay! by jimtheowl · · Score: 2

      You do not need to. Free Pascal is well and alive on multiple platforms and is open source. http://www.freepascal.org/

  4. Can this thing by Anonymous Coward · · Score: 0

    Run Doom maxed out or nah?

    1. Re:Can this thing by Anonymous Coward · · Score: 0

      Anything can run Doom nowadays, even my MP3 player

    2. Re: Can this thing by Anonymous Coward · · Score: 0

      You still own an mp3 player, and it's a rockbox, of all things? Was this posted from 2002?

    3. Re: Can this thing by Anonymous Coward · · Score: 0

      My MP3 player (a SanDisk Zip Clip) is tiny, maybe the size of a matchbox, it holds a ton of music, it doesn't run down the battery on my phone and it's much less annoying to carry when I exercise or just when walking around. I'd also rather lose or damage the player than my phone when doing said exercises. Even Apple still sells iPods.

      Rockbox is the best interface and player firmware there is. I think it's sad that in 15 years, nobody else could surpass it.

      That said, I am sorry that you are so out of touch with real life. Maybe you should take a break from your computer and get out more often.

  5. Tsubame-KFC by Anonymous Coward · · Score: 0

    Had no idea KFC was active in the AI field. Good Job, Kernel Sanders.

    1. Re:Tsubame-KFC by Anonymous Coward · · Score: 2, Informative

      Kepler Fluid Cooled, though Tsubame is a kind of bird and the computers are indeed immersed in oil.

      Also Japan has an unusual tradition of eating KFC for Christmas with reservations made months in advance.

    2. Re:Tsubame-KFC by cant_get_a_good_nick · · Score: 1

      For some reason, Kernel Sanders made me think of Dr Fun: Kernel Panic

      For those that don't know, Dr Fun was the first webcomic 520 weeks, or 10 years worth

  6. Not alone by SuperKendall · · Score: 2

    That is actually the first thing that sprang to mind, even though I had been looking specifically at Pascal based GPU's recently. :-)

    --
    "There is more worth loving than we have strength to love." - Brian Jay Stanley
    1. Re:Not alone by gander666 · · Score: 1

      Ditto here. I had some memories of long dead undergrad programming courses.

      --
      Suppose you were an idiot and suppose you were a member of Congress ... but I repeat myself. - Mark T
    2. Re:Not alone by Anonymous Coward · · Score: 0

      In terms of bulk SLOC, pascal is my #1 career language. Even though I haven't touched it in a decade.

    3. Re:Not alone by Anonymous Coward · · Score: 1

      I'm older school and keep hoping for an ALGOL supercomputer.

    4. Re:Not alone by Applehu+Akbar · · Score: 2

      "Ditto here. I had some memories of long dead undergrad programming courses."

      Yes, Pascal, really? If we're going to invoke the age of steam, why not go full Fortran with this project?

    5. Re:Not alone by Anonymous Coward · · Score: 0

      Latin is older than English, but a thousand times better as a language.

    6. Re:Not alone by JanneM · · Score: 1

      Fortran is still very much in use for mathematical programming.

      --
      Trust the Computer. The Computer is your friend.
    7. Re: Not alone by Anonymous Coward · · Score: 0

      Hard language is English

    8. Re:Not alone by Darinbob · · Score: 2

      Makes no sense. Pascal is a perfectly fine language, even classic Pascal. Fortran, especially classic Fortran 77, is really lacking in many ways (newer versions add newer features but that's like comparing Visual Basic to Basic and trying to call them the same language). Granted Modula-II or Ada is much more suitable than Pascal.

    9. Re:Not alone by Applehu+Akbar · · Score: 1

      Latin is older than English, but a thousand times better as a language.

      The value of Latin is not that it's "better" in modern usage but that as a root of so many later European languages, it is easily used as a lingua franca for European speakers. An Italian MD, a British MD and a French MD all understand what bis in diem on a prescription vial means.

    10. Re:Not alone by JustAnotherOldGuy · · Score: 1

      Latin is older than English, but a thousand times better as a language.

      Which must be why it's used so today. Oh, wait....

      --
      Just cruising through this digital world at 33 1/3 rpm...
    11. Re:Not alone by JustAnotherOldGuy · · Score: 2

      "English is a language that lurks in dark alleys, beats up other languages and rifles through their pockets for spare vocabulary."

      --attributed to James D. Nicoll

      The original, complete quote appears to be:

      "The problem with defending the purity of the English language is that English is about as pure as a cribhouse whore. We don’t just borrow words; on occasion, English has pursued other languages down alleyways to beat them unconscious and rifle their pockets for new vocabulary."

      --
      Just cruising through this digital world at 33 1/3 rpm...
    12. Re:Not alone by Anonymous Coward · · Score: 1

      Your position might be viewed as an example of fallacy of the single cause with elements of argumentum ad populum. -PCP

    13. Re:Not alone by paai · · Score: 1

      Perhaps APL should be revived. Imagine the money to be made by keyboard manufacturers.

      Paai

    14. Re:Not alone by mikael · · Score: 1

      I'm worried the animal activists might take offence at peta-flops.

      --
      Vintage computer adverts: http://www.vintageadbrowser.com/computers-and-software-ads
    15. Re:Not alone by Anne+Thwacks · · Score: 1
      ALGOL supercomputer.

      Now we have multi-threading, Algol68 might actually be the best choice. However, history supports C-more Cray.

      --
      Sent from my ASR33 using ASCII
    16. Re: Not alone by Anonymous Coward · · Score: 0

      And a Burroughs stack machine to go with it..

    17. Re:Not alone by Godwin+O'Hitler · · Score: 1

      Yes, I too would rather have a language with a million inflections to save a dozen pronouns and structured so inflexibly that you have to add a zillion articles and conjunctions.

      --
      No, your children are not the special ones. Nor are your pets.
    18. Re:Not alone by RespekMyAthorati · · Score: 1

      Fortran, especially classic Fortran 77, is really lacking in many ways

      But perfectly adequate for solving differential equations, which is the main task of scientific and engineering programs.

    19. Re:Not alone by michael_wojcik · · Score: 1

      APL was effectively superseded by the J language.

      I taught myself some APL[1] using a standard keyboard and an APL implementation that ran on OS/2, if memory serves. Wasn't bad. Yes, you had to compose all the APL characters using Alt-key sequences[2], but APL is so terse that you didn't have to do it very often.

      J uses ASCII rather than APL's grab-bag of symbols. I'm not sure I'd call it more readable, though.

      It would be interesting to see a less-opaque Iverson-Backus language, with the matrix operators and tacit-programming features.

      [1] Because I believe it's useful to try as many programming-language paradigms as possible, and because I'd read Iverson's and Backus' Turing Award addresses.

      [2] Can't remember whether Alt-Right gave you the GoTo operator or the NarcissisticPersonalityDisorder operator.

    20. Re:Not alone by painandgreed · · Score: 1

      Latin is older than English, but a thousand times better as a language.

      Which must be why it's used so today. Oh, wait....

      Saw an article on a study once about the value of either language standardization and regularness versus freeform and irregular usage given between the examples of French and English. It seems that the irregular and modifyable nature of English actually helps more towards usage and adoption than standardization.

  7. Good ol' days by Tablizer · · Score: 0

    I used to use Pascal a lot back in my college days, mostly on mini-computers. Other than a brief burst of sales in Turbo Pascal (PC) and to a lessor extent Delphi, Pascal usage quickly shrank. I'm not quite sure why, it was a fairly decent compiler-based language.

    It needed more string-oriented operations, perhaps. I like the way the type name (declaration) comes after the variable, instead of before like the C-family of languages. I prefer it after. It also allowed nesting of functions.

    1. Re:Good ol' days by Anonymous Coward · · Score: 0

      Just think, with this you can have the world's fastest Karel the Robot.

    2. Re:Good ol' days by MightyMartian · · Score: 1

      I had a pretty deep fondness for Pascal back in the day, and messed around with Delphi, Modula and Oberon, but the reality is that these aren't exactly common languages anymore, at least not in commercial circles. It's a real pity too, because learning TurboPascal was my sort of "Wizard of Oz black-and-white to color" moment back in highschool, where I shed all the evils that I had learned through mucking around with various flavors of BASIC, and basked in the glory of structured procedural code.

      --
      The world's burning. Moped Jesus spotted on I50. Details at 11.
    3. Re:Good ol' days by Aighearach · · Score: 1

      I thought all Pascal use was for BBS DOOR games, and the internet killed it?

    4. Re:Good ol' days by cheesybagel · · Score: 2

      Pascal usage quickly shrank. I'm not quite sure why, it was a fairly decent compiler-based language.

      I am pretty sure why. More modern operating systems came up and nearly invariably they were all C based. If you wanted to call the OS libraries you were better off using C. That's why Pascal died off.

    5. Re:Good ol' days by Anonymous Coward · · Score: 0

      > I thought all Pascal use was for BBS DOOR games, and the internet killed it?

      What killed Pascal was the desperate need to squeeze performance out of those old chips of yore: 8086, 8088, 80286, etc. People were forced to use C, as it ran closer to the hardware and allowed some extra optimizing.

      Ironically, there's much less need for it these days. People abuse the use of interpreters -- even inside browsers -- and Pascal would be fast enough for modern applications, IMHO. C even is experiencing a reduction in popularity, if surveys like Tiobe are to be believed.

    6. Re:Good ol' days by phantomfive · · Score: 1

      It also allowed nesting of functions.

      FWIW GCC allows that now in C.

      --
      "First they came for the slanderers and i said nothing."
    7. Re:Good ol' days by paai · · Score: 1

      Faster hardware is not always a boon. Now we have Java monsters that eat up all the performance of even decent hardware. Easy to learn? Nah, with the myriads of different libraries and paths it is a conceptual mess. And it was supposed to be the cure-all for viruses, but that has not materialized either.

      The good Lord gave us C, Bash and the CLI , and we frail humans should not presume to improve on His creation.

      Paai

    8. Re: Good ol' days by dunkelfalke · · Score: 1

      It is good that I am an atheist, then.

      --
      "It's such a fine line between stupid and clever" -- David St. Hubbins, Spinal Tap
    9. Re:Good ol' days by jbolden · · Score: 0

      The reason is both the reason for its success and the reason for its failure. The Pascal language makes a lot of compromises in areas of readability and organization to allow for small compilers. In the case of PCs, it was much easier to write a Pascal compiler that ran well off a 128k floppy than a C compiler. That stopped mattering pretty quickly.

    10. Re:Good ol' days by angel'o'sphere · · Score: 1

      You can call a C library or OS functiom from Pascal, too.
      There is no difference.

      C is not something magically, you know ...

      --
      Cost free eBook I read (by iBook/Kobo/Amazon/ObookO/Gutenberg etc.): "The Green Odyssey" by Philip Jose Farmer.
    11. Re:Good ol' days by haruchai · · Score: 1

      Faster hardware is not always a boon. Now we have Java monsters that eat up all the performance of even decent hardware. Easy to learn? Nah, with the myriads of different libraries and paths it is a conceptual mess. And it was supposed to be the cure-all for viruses, but that has not materialized either.

      The good Lord gave us C, Bash and the CLI , and we frail humans should not presume to improve on His creation.

      Paai

      Those are tools of the Devil; on the 1st day, the Lord created binary, microcode and logic gates and saw that it was good.

      --
      Pain is merely failure leaving the body
    12. Re:Good ol' days by cheesybagel · · Score: 1

      I don't know if there is any difference between calling conventions, but the mere fact you would have to write your own headers in Pascal seems pretty significant to me.

    13. Re:Good ol' days by angel'o'sphere · · Score: 1

      Ofc, there are different calling conventions.
      Usually C pushes arguments from left to right onto the stack and the caller cleans up the stack (because of variable argument lists), in Pascal arguments get pushed from right to left and the called procedure/fuction cleans up the stack.

      Interfacing with C you usually do via so called 'units'. Units have an interface section and an implementation section. In the interface section you define functions/procedures and call also define if they are written in a different language (Assembly, C, Fortran), unfortunately there is no standard how to do that exactly. (Implementation section would be empty and you have to link with the relevant C library, ofc.)

      Most C compilers also used to support 'extern PASCAL' or 'extern FORTRAN' keywords/declarations. But again I think there is no standard for that.

      --
      Cost free eBook I read (by iBook/Kobo/Amazon/ObookO/Gutenberg etc.): "The Green Odyssey" by Philip Jose Farmer.
  8. Good for them... by Arkh89 · · Score: 1

    These P100 come with sweet HBM2 and around 500GB/s in memory bandwidth... everything based on dense linear algebra (AI, but also physics simulations) is basically flying on them.

    1. Re: Good for them... by Anonymous Coward · · Score: 0

      Do they run their Tesla P100 GPUs in Ludicrous mode?

  9. Extra Crispy by fyngyrz · · Score: 2

    experimental immersion-cooled Tsubame-KFC system

    I'm sorry, but I like my KFC hot from the fryer. This experiment should end immediately.

    --
    I've fallen off your lawn, and I can't get up.
    1. Re:Extra Crispy by Anonymous Coward · · Score: 0

      Typical meat-eater demonstrates stupidity due to meat-addled brain, film at -11.

    2. Re:Extra Crispy by Anonymous Coward · · Score: 0

      More like due to grease-addled brain. KFC and fried foods in general are fucking horrible.

    3. Re:Extra Crispy by fyngyrz · · Score: 1

      KFC and fried foods in general are fucking horrible.

      Well, there's your problem, right there. Why not try the appropriate product which was meant for that instead?

      No, no need to thank me. Where would the world be if us experienced types didn't pass along our wisdom to the clueless?

      --
      I've fallen off your lawn, and I can't get up.
  10. Oh Yeah by Anonymous Coward · · Score: 0

    I am going to make make a supercomputer that runs on BASIC A.

    1. Re:Oh Yeah by Crashmarik · · Score: 2

      I am going to make make a supercomputer that runs on BASIC A.

      Whom the gods would destroy, first they teach basic.

    2. Re:Oh Yeah by hackwrench · · Score: 1

      And the first step to destroy the gods is to learn BASIC. The next step is to learn how to create and embed your own commands into a BASIC compiler. But tread carefully! The believers in the gods seek out those who would destroy the gods. A man named Steve McNeill, http://www.qb64.net/forum/inde... started down that path and was brought low, and a recent accounting of his tale was scrubbed from the annals of http://www.qb64.net/forum/inde...

  11. Worst. Name. Ever. by Anonymous Coward · · Score: 0

    What were they thinking? I guess the product is better than the name if it's gaining traction, but for the love of God, fix the name. There's one thing in computing that's Pascal, and these chips are not it.

  12. Half Precision by Anonymous Coward · · Score: 0

    Of course Kepler should support it with software, at the same speed as single precision. The selected implementations of Pascal brought the speedier half precision support to the table.

  13. This = Pascal based: Object Pascal Delphi by Anonymous Coward · · Score: 0, Redundant

    APK Hosts File Engine 9.0++ SR-7 32/64-bit https://www.google.com/search?hl=en&source=hp&biw=&bih=&q=%22APK+Hosts+File+Engine%22+and+%22start64%22&btnG=Google+Search&gbv=1/

    Ads/malware rob speed/security/privacy

    Hosts add speed (via hardcodes/adblocks), security (vs. bad sites/malware/poisoned dns), reliability (vs. dns down), & anonymity (vs. dns requestlogs/trackers).

    Less power/cpu/ram + IO vs. DNS/routers/addons/antivirus + less security bugs/complexity & faster vs. addons/routers/remote dns!

    * Via what you NATIVELY have built in the IP stack in FASTER kernelmode!

    Hosted/recommended by Malwarebytes' hpHosts + code verified safe by 'em.

    APK

    P.S. - Safe https://www.virustotal.com/en/file/e01211ca36aa02e923f20adee0a3c4f5d5187dc65bdf1c997b3da3c2b0745425/analysis/1433430542/ & No security issue/bug in it since 2012 release & virus proof HyperAlloy Combat Chassis - Microprocessor controlled: Fully Armored, VERY tough code construction... apk

  14. Luckily.. by Anonymous Coward · · Score: 0

    our product which is celebrating it's 30th anniversary, is written in Delphi. Now, it will finally be able to understand what it's purpose in the world is.

  15. RTFA by subk · · Score: 3, Insightful

    Pascal, the GPU design. Not Pascal, the language.

    --
    Now, if you'll excuse me, I have backups to corrupt.
    1. Re:RTFA by Hognoxious · · Score: 2

      It does make one wonder what they were thinking - if they were at all - when they chose the name.

      --
      Confucius say, "Find worm in apple - bad. Find half a worm - worse."
    2. Re:RTFA by Walter+White · · Score: 2

      Pascal, the GPU design. Not Pascal, the language.

      So... It won't run Delphi then?

    3. Re:RTFA by HornWumpus · · Score: 0

      Delphi closed, along with BIX, Prodogy and all the other AOL like online services of the early 90s.

      --
      John McAfee 'It was like that time I hired that Bangkok prostitute; to do my taxes, while I fucked my accountant'
    4. Re:RTFA by Anonymous Coward · · Score: 0

      If not Delphi, what about Free Pascal?

    5. Re:RTFA by Anonymous Coward · · Score: 0

      Thanks - saved me reading TFA to answer the obvious question :)

    6. Re:RTFA by Walter+White · · Score: 1

      Delphi closed, along with BIX, Prodogy and all the other AOL like online services of the early 90s.

      They're all gone? Compuserve too? Sad!

    7. Re:RTFA by Anonymous Coward · · Score: 0

      They were probably thinking of Blaise Pascal, considering his well-known work in the field of projective geometry. -PCP

    8. Re:RTFA by Hognoxious · · Score: 1

      Probably true, but irrelevant. The point is that they didn't think to check if the word was already in use in a related context.

      --
      Confucius say, "Find worm in apple - bad. Find half a worm - worse."
  16. world's largest supplier of quiche by Anonymous Coward · · Score: 0

    to open in japan :)

  17. One of few people by TheOuterLinux · · Score: 1

    I think I'm one of few people that actual likes Pascal. Also prefer Python over JAVA and never really cared for C all that much even though there are similarities. Anyone like or use MyNotex (Linux)? Written in with Pascal. ;)

  18. Re:Worst. Name. Ever. by Anonymous Coward · · Score: 0

    What were they thinking? I guess the product is better than the name if it's gaining traction, but for the love of God, fix the name. There's one thing in computing that's Pascal, and these chips are not it.

    Here at Nvidia we unfortunately suffer from marketing sometimes being lazy and by default naming some products the same as our internal project code names to avoid being confusing.

    FWIW, Blaise Pascal's first published written work was called, Essay on Conic Sections. The writings constituted an important advance in projective geometry, (representing a 3-D object onto a 2-D field). Perhaps a reasonable internal codename for a 3d graphics chip project? Arguable. A name collision? Unfortunate. Are concept name collisions uncommon with prolific scientists? No. For example Galios, Gauss, Euler, etc

    It's not like computing doesn't do stuff like that all the time. Witness JavaScript and C-shell... JavaScript is definitely not Java and C-shell is definitely not C.

  19. GPUs have limited applications by 140Mandak262Jamuna · · Score: 3, Insightful
    It keeps coming back. Massively parallel machine, thousands of cores all working in parallel. Naively multiply add up all the megaflops and get some massive number and tout it big. We can simply add all the flops of all the servers in some Amazon cloud and claim that is the super computer. Back in the 80s "transputer" was all the rage. Before that it was the "vector" computers. Then "the network *is* the computer", then GPUs...

    As of now there are very few applications for massively souped up GPU processes. Fluid mechanics loves this GPU. Navier-Stokes is probably the most difficult equation to solve, agreed. But it is hyperbolic, with limited "zone of influence", and numerical equations are quite simple, just mass, momentum and energy balances in the control volume. It plays well in GPU, the calculations fit inside the teensy memory and processor. All time domain problems are hyperbolic and they all can be ported to GPU, theoretically. But try squeezing Maxwell's Equations into that teensy processor!

    Graphics card companies are desperately looking for new markets and they keep pushing this. They might as well push a wet noodle across the table. It ain't gonna go nowhere it didn't wanna go.

    --
    sed -e 's/Chuck Norris/Rajnikant/g' joke > fact
    1. Re:GPUs have limited applications by cheesybagel · · Score: 1

      I wouldn't call a modern GPU that constrained really. The only thing they lack is memory protection. It's also a lot easier to program on a GPU ever since the SIMT paradigm came out (i.e. CUDA, OpenCL). Also plenty of modern processors come with a GPU on the same die as the CPU. Like nearly all smartphones for example.

    2. Re:GPUs have limited applications by Anonymous Coward · · Score: 0

      Though there's some progress on this front, GPUs are problematic because there's not a good way to compute matrix factorizations on them. In fact, until cuBLAS-XT came out, there wasn't even a convenient, fast way to do matrix-matrix multiplication on NVIDIA GPUs. GPUs are really good at solving PDEs that can be handled with explicit time stepping methods. There are plenty of problems that require implicit methods and they don't benefit from GPUs. I'll also reiterate that the factorization issue is a big one. Matrix factorizations aren't just mathematical constructs, but physical. For example, the polar factorization divides a linear transformation into a stretch plus a rotation, or vice versa. This is intimately related to things like the SVD. Until these problems can easily, quickly, and reliably be solved on GPUs, then they are more limited than conventional methods, which means using MPI on a cluster with a fast interconnect.

    3. Re:GPUs have limited applications by slew · · Score: 1

      I wouldn't call a modern GPU that constrained really. The only thing they lack is memory protection. It's also a lot easier to program on a GPU ever since the SIMT paradigm came out (i.e. CUDA, OpenCL). Also plenty of modern processors come with a GPU on the same die as the CPU. Like nearly all smartphones for example.

      Actually all modern GPU have had memory protection for several generations. The problem GPUs have is that don't generally have full support for demand paging and precise exceptions (SIMT makes that pretty expensive). For example, putting in hardware for highest possible performance and hardware to be able to hit a page fault and be able to clean up and restart multiple threads (that might be communicating or synchronizing states) are two different hardware optimization points. That being said, some limited support for demand paging is generally possible.

    4. Re:GPUs have limited applications by slew · · Score: 1

      GPUs are problematic because there's not a good way to compute matrix factorizations on them.

      Have you looked at this paper on large matrix factorization?

    5. Re:GPUs have limited applications by religionofpeas · · Score: 1

      This one is targeted towards neural nets. Does that work ?

    6. Re:GPUs have limited applications by Anonymous Coward · · Score: 0

      That's interesting. Thanks! What they're doing there is a low rank factorization, which is sometimes helpful for scientific problems, but not as often. Really, we need an LU, Choleski, QR, SVD, and spectral decomposition for dense matrices as a start. Honestly, as much as it's maligned, the routines in LAPACK are really what applied mathematicians require for solving a broad class of PDEs with a broad class of algorithms. Certainly, we need sparse routines as well, but these dense routines are often used in the sparse algorithms. For example, supernodal and multifrontal algorithms for sparse Choleski take advantage of dense operations. Even preconditioners like algebraic multigrid typically have a factorization at their inner most layer.

      Long story short, a library like cuBLAS-XT needs to be extended to all of level 3 BLAS routines. We need the XT version because asking a programmer to partition the matrix on their own and stitch it back together is completely unacceptable. Then, we essentially need a cuLAPACK-XT library. Yes, useful problems can be solved prior to then. However, until that point, when someone asks whether GPUs are more limited in an application setting, the answer is a resounding yes and we can easily point to applications that depend on these missing routines.

    7. Re:GPUs have limited applications by Anonymous Coward · · Score: 0

      We need the XT version because asking a programmer to partition the matrix on their own and stitch it back together is completely unacceptable.

      We need this statement written in Golden Letters at the entrance of... say, every school, subway station, highway toll, airports, fast food chain stores, hotspots, cinemas and important historical monument? That would do.

    8. Re:GPUs have limited applications by 140Mandak262Jamuna · · Score: 1

      I don't know neural networks. Computational physics, applied math background.

      --
      sed -e 's/Chuck Norris/Rajnikant/g' joke > fact
    9. Re:GPUs have limited applications by Anonymous Coward · · Score: 0

      Certainly, we need sparse routines as well, but these dense routines are often used in the sparse algorithms.

      I'm not an expert but I really thought that the sparse is the real issue with the most current GPUs as dense problems are more suitable for the vector machine. Partition problems are surely performance problems as well, with the platform dependencies and performance portability nightmares. As bandwidth increases and the CPU/GPU logical distance shortens, perhaps thighs will get easier in the future. Commercial and custom software from smaller shops really need platform stability if the libraries are not there.

    10. Re:GPUs have limited applications by Anonymous Coward · · Score: 0

      You know what they are really good at. The specific thing this machine was built for. Like basically every super computer before it. It was design and built for a specific set of problems. Not all problems. Oh and Maxwell problems are also local (In all possible frames of reference and are EVEN linear unlike NS). In fact that was how solvers for them worked 20 years ago, so no idea what your going on about.

    11. Re:GPUs have limited applications by 140Mandak262Jamuna · · Score: 1

      Problem is Maxwell is the null space of the curl vector. It results in a rank deficient matrix, without unique solution. It wasn't until 1980 they figured out a way to subtract the null space and uniquify the solution. But that matrix does not have diagonal dominance. So all iterative methods are out. The condition number is very poor. LU decomposition is the only way out. And the unknowns are complex vectors. Not simple scalars like density, pressure and one real velocity vector for NS. Squeezing it all inside a GPU processor is very difficult.

      --
      sed -e 's/Chuck Norris/Rajnikant/g' joke > fact
  20. Pascal? by hublan · · Score: 1

    Making all the other supercomputers Wirth-less.

    --
    My spoon is too big.
    1. Re:Pascal? by slew · · Score: 1

      Making all the other supercomputers Wirth-less.

      Or by name, Veert-ually worthless...

  21. Re:Good ol' days - blame p2c by Anonymous Coward · · Score: 1

    One it was realized you could mechanically translate Pascal to C, then compile the C and get a 2-3x speed up (compared to interpreting p-code), Pascal started dying.

    C has issues, but in practice it has less of them than Pascal, so the resulting code was easy to tweak to get even more performance.

  22. Borland by s.petry · · Score: 2

    Delphi.. Ohhh such wonderful memories..

    --

    -The wise argue that there are few absolutes, the fool argues that there are no probabilities.

    1. Re:Borland by haruchai · · Score: 1

      Delphi.. Ohhh such wonderful memories..

      I've heard its latest incarnation as Embarcadero Delphi is quite good but I'm no judge of programming environments

      --
      Pain is merely failure leaving the body
  23. Re:Chicken vs. Swallow by hackwrench · · Score: 1

    And you probably prefer your KFC to be chicken and not swallow, which is what Tsubame means.

  24. Re: Worst. Name. Ever. by Anonymous Coward · · Score: 0

    Tesla, Fermi, Kepler, Maxwell, Pascal, Volta; who's next?

  25. Re: Worst. Name. Ever. by Anonymous Coward · · Score: 0

    Tesla, Fermi, Kepler, Maxwell, Pascal, Volta; who's next?

    If you can find someone who *broke the code*, you will be rewarded with the answer.
    I could tell you, but then you'd be pushing up daisies too ;^)

  26. At least it's not MS-BASIC!!! by Anonymous Coward · · Score: 0

    10 REM Jeez!! Imagine if you had to program an artificial intelligence using line numbers and goto statements.

  27. Re:Chicken vs. Swallow by fyngyrz · · Score: 1

    Precisely so. Well played. :)

    --
    I've fallen off your lawn, and I can't get up.
  28. Japanese Engineers Strike Again by Anonymous Coward · · Score: 0

    To name a computer system, Tsubame-KFC, after Kentucky Fried Chicken, the experimental immersion-cooled system and Tsubame "Swallow" maybe not the Bird but the action verb, is Genius.

    Ja ja

    1. Re:Japanese Engineers Strike Again by Anonymous Coward · · Score: 0

      Makes me think some KFC owner is the perpetrator of the thingie and Japanese is just an excuse to avoid journalists and have cheap foreign hand labor costs...

  29. Re:Good ol' days - blame p2c by Anonymous Coward · · Score: 0

    > One it was realized you could mechanically translate Pascal to C, then compile the C and get a 2-3x speed up (compared to interpreting p-code), Pascal started dying.

    The gains were not so great. After all, p-code was gradually replaced by very efficient compilers (like Turbo Pascal and later, Delphi). It would be interesting to look at some more recent benchmarks...

    There was that emphasis on compile performance and runtime speed was less of a worry; with C, it was the opposite, compilation was much slower, but the final executable could be finely tuned -- like you say in "the resulting code was easy to tweak to get even more performance". This was absolutely fundamental for games, graphical programs, CAD and other CPU-intensive tasks.

    And by trying to be good at tasks where performance was less important, Pascal had also to face Java as a competitor.