Slashdot Mirror


IBM Building 20 Petaflop Computer For the US Gov't

eldavojohn writes "When it's built, 'Sequoia' will outshine every super computer on the top 500 list today. The specs on this 96 rack beast are a bit hard to comprehend as it consists of 1.6 million processors and some 1.6TB of memory. That's 1.6 million processors — not cores. Its purpose? Primarily to keep track of nuclear waste & simulate explosions of nuclear munitions, but also for research into astronomy, energy, the human genome, and climate change. Hopefully the government uses this magnificent tool wisely when it gets it in 2012."

248 comments

  1. and just for old time's sake... by Hell+O'World · · Score: 4, Funny

    Can you imagine a Beowolf cluster of those?

    1. Re:and just for old time's sake... by skund · · Score: 1

      Wich framerate will it get in quake?

    2. Re:and just for old time's sake... by Anonymous Coward · · Score: 5, Funny

      Can you imagine a Beowolf cluster of those?

      No. No I can't. I can't imagine a beowolf cluster of one those. Even if Natalie Portman (covered in grits) was in my base killing my overlords like an insensitive clod. Even if netcraft confirmed it, then it confirmed netcraft in soviet russia. Especially if Cowboy Neal gave me a three step plan leading to profit I could not imagine it.

      Enough with the meme. Or not, because I must be new here.

    3. Re:and just for old time's sake... by Anonymous Coward · · Score: 1, Funny

      you mean Crysis

    4. Re:and just for old time's sake... by Anonymous Coward · · Score: 0

      Pics or it didn't happen

    5. Re:and just for old time's sake... by Anonymous Coward · · Score: 5, Interesting

      While you are joking about games (like Quake and Crysis), this computer does sound like a giant graphics card.

      It can do 20 Pflops with 1.6 million processors, so 12.5Gflops per processor, but with 1.6TB of memory, it means its only got 1Mb per processor.

      So it sounds like some kind of giant specialised GPU with local memory.

    6. Re:and just for old time's sake... by Anonymous Coward · · Score: 0

      Okay, this is way off topic. But I was reading a story about the Italians putting Google executives on trial for a youtube video. I refresh and the story was gone. Anyone know what happened?

    7. Re:and just for old time's sake... by Anonymous Coward · · Score: 1, Interesting

      Sounds like the Cell on steroids to me.

    8. Re:and just for old time's sake... by Anonymous Coward · · Score: 0

      Glad I saw your post while I still my other tab open. In case anyone's wondering, he's talking about this story:

      Google Execs Being Tried In Italy Over User's "Cruelty" Video
      Posted by timothy on Tuesday February 03, @10:12AM
      from the what-have-the-googlers-ever-done-for-us? dept.
      netbuzz writes "Italian officials this morning have begun prosecuting four Google executives over their roles -- which were essentially non-existent -- in the posting of a video that depicted cruelty toward a disabled child. "It's akin to prosecuting mail service employees for hate speech letters sent in the post," the spokeswoman said in an e-mail. "Seeking to hold neutral platforms liable for content posted on them is a direct attack on a free, open Internet."

    9. Re:and just for old time's sake... by Sfing_ter · · Score: 1

      It's actually going to be serving QuakeLive when it goes live ... finally ...

      --
      A computer once beat me at chess, but it was no match for me at kick boxing. Emo Philips
    10. Re:and just for old time's sake... by ed.mps · · Score: 5, Funny

      damn you AC, this post deserves a +50 "beowulf cluster of memes"

      --
      !sig
    11. Re:and just for old time's sake... by Anonymous Coward · · Score: 0

      "sounds like bad math to me"

      LOL!, you just described your own post!

      1.6TB / 1.6M = 1MB per CPU

      (And to check, as each CPU has 1MB and there are 1600000 CPUs, then its a total memory of 1600000 * 1MB = 1600000MB = 1600GB = 1.6TB)

    12. Re:and just for old time's sake... by Anonymous Coward · · Score: 0

      Yeah, whatever. But does it run linux?

    13. Re:and just for old time's sake... by Forge · · Score: 1

      It probably dose. Other IBM Built SCs do.

      --
      --= Isn't it surprising how badly I spell ?
    14. Re:and just for old time's sake... by Anonymous Coward · · Score: 1, Funny

      It will run Duke Nukem Forever on Vista!

    15. Re:and just for old time's sake... by fataugie · · Score: 1

      You forgot goat.cx

      Unfortunatly, I can't...

      --

      WTF? Over?

    16. Re:and just for old time's sake... by fataugie · · Score: 1

      I misspelled it...my bad

      goatse.cx

      --

      WTF? Over?

    17. Re:and just for old time's sake... by hydromike2 · · Score: 1

      dont forget cores, sounds like they will have more L2 cache then actual ram

    18. Re:and just for old time's sake... by RabidMoose · · Score: 1

      I knew it! They're building the cell phone sonar-imaging device from Dark Knight!

    19. Re:and just for old time's sake... by Anonymous Coward · · Score: 0

      Congratulations, you just broke the internet. I hope you're happy.

    20. Re:and just for old time's sake... by lordsid · · Score: 1

      No conspiracy, it has merely faded to yesterday's news. Which can be found in the right column.

      --
      IMAGE VERIFICATION IS EVIL!
    21. Re:and just for old time's sake... by clodney · · Score: 2, Informative

      The summary is wrong. I actually did RTFA, and it said 1.6 petabytes, not 1.6 terabytes.

    22. Re:and just for old time's sake... by Anonymous Coward · · Score: 0

      this assumes each processor is running in tandem at all times. if only 25% of it's processing power and 45% of it's memory capacity was used to complete one task while the remaining resources are divided up among other task then your 1MB theory falls a part.

      as for falling apart goes we may now no our fate. the Mayan calender is scheduled to end about the same time our government gets there new toy.

    23. Re:and just for old time's sake... by DMUTPeregrine · · Score: 2, Informative

      IIRC it switched to goat.cx for a while. So both are correct. Also, be glad you haven't seen the stereogram of it. Some things should not be in 3d.

      --
      Not a sentence!
    24. Re:and just for old time's sake... by porky_pig_jr · · Score: 2, Funny

      In Soviet Russia a beowulf cluster of those imagines you.

    25. Re:and just for old time's sake... by Kayden · · Score: 1

      The Mayans predicted Skynet...

    26. Re:and just for old time's sake... by religious+freak · · Score: 1

      I, for one, welcome my sarcastic AC overlord's comment

      --
      If you can read this... 01110101 01110010 00100000 01100001 00100000 01100111 01100101 01100101 01101011
    27. Re:and just for old time's sake... by HTH+NE1 · · Score: 1

      It can do 20 Pflops with 1.6 million processors, so 12.5Gflops per processor, but with 1.6TB of memory, it means its only got 1Mb per processor.

      The other seven megabits per processor are used for parity checking.

      --
      Oh, say does that Star-Spangled Banner entwine / The myrtle of Venus with Bacchus's vine?
    28. Re:and just for old time's sake... by tellthepeople · · Score: 2, Informative

      Can you imagine a Beowolf cluster of those?

      I don't think that phrase means what he thinks it means. http://en.wikipedia.org/wiki/Beowulf_(computing)

      p.s It's Beowulf not Beowolf.

      --
      Tanto nomini nullum par elogium.
    29. Re:and just for old time's sake... by Anonymous Coward · · Score: 0

      I don't get it. Could you explain with a car analogy?

    30. Re:and just for old time's sake... by HTH+NE1 · · Score: 2, Informative

      Except of course that TFA was wrong and it was 1.6 PB / 1.6 MCPUs = 1 GB/CPU.

      I imagine though it could run a few gigs (jobs of short or uncertain duration).

      --
      Oh, say does that Star-Spangled Banner entwine / The myrtle of Venus with Bacchus's vine?
    31. Re:and just for old time's sake... by howlingmadhowie · · Score: 1

      interestingly this is more than the amount of memory the mmu for the athlon 64 can manage, it being limited to 48 significant bits in the 64 bit address space.

    32. Re:and just for old time's sake... by Hurricane78 · · Score: 1

      My Qualculate says

      (1,6TiB / 1,6 million) to MiB = (1,6 tebibyte) (1,6 million) to mebibyte
      = 1,048576 MiB

      --
      Any sufficiently advanced intelligence is indistinguishable from stupidity.
    33. Re:and just for old time's sake... by SpacePunk · · Score: 1

      I think it's goatse.cz now.

      I linked it on ATS forums as the result of alien contact. They were not amused.

    34. Re:and just for old time's sake... by H3g3m0n · · Score: 1

      Personally I think its some kind of secret AI SkyNet style research project. The current petaflop computers are fairly close to having the same computing capacity as the human brain based on best estimates but thats not the same as having the computing power to simulate one. This system is 20x more powerful though...

      --
      cat /dev/urandom > .sig
    35. Re:and just for old time's sake... by turing_m · · Score: 1

      Especially if Cowboy Neal gave me a three step plan leading to profit I could not imagine it.

      It looks like you are having trouble imagining a beowulf cluster. Perhaps a car analogy would help?

      --
      If I have seen further it is by stealing the Intellectual Property of giants.
    36. Re:and just for old time's sake... by Anonymous Coward · · Score: 0

      Or even in 2d.

  2. Mmm... by Taibhsear · · Score: 5, Funny

    Nice rack(s).

  3. OH NOES!!! by Darundal · · Score: 3, Funny

    2012! Supercomputer! It's Skynet! RUN FOR YOUR LIVES!

    1. Re:OH NOES!!! by GiovanniZero · · Score: 5, Funny

      At 6:18 pm EST IBM's super computer went online At 6:19 pm EST IBM's super computer declared nuclear war on humans At 6:20 pm EST there was a SEG FAULT and skynet must reboot to continue genocide

      --
      Mod me up, mod me down, do your worst you modding clown.
    2. Re:OH NOES!!! by need4mospd · · Score: 1

      Skynet? More like Multivac.

    3. Re:OH NOES!!! by CompMD · · Score: 2, Informative

      Well, if IBM builds skynet, then we win the war by saying PWRDNSYS OPTION(*IMMED).

    4. Re:OH NOES!!! by Anonymous Coward · · Score: 0

      Probably a UAC prompt:

      "The following program needs your permission to continue nuclear holocaust:
      Skynet
      Would you like to allow it?"

      BSOD!

    5. Re:OH NOES!!! by blhack · · Score: 1

      At 6:20 pm EST there was a SEG FAULT and skynet must reboot to continue genocide

      I think what you mean to say was:

      Operatior action required on device SKYNET1A (Cancel Reply Ignore)
      Unchecked Function

      Reply:__________________________________________________

      --
      NewslilySocial News. No lolcats allowed.
    6. Re:OH NOES!!! by carlmenezes · · Score: 1

      At 6:22 pm EST skynet does a self diagnosis of the SEG FAULT and attacks Microsoft.
      At 7:00 pm EST all online Windows computers start self-installing Linux.

      --
      Find a job you like and you will never work a day in your life.
    7. Re:OH NOES!!! by dwywit · · Score: 1
      Ah, the joys of OS/400. I remember those days, playing with the key and switches of the service processor on the front of a 9406 rack....

      I wonder if this monster will have something similar - IPL from the 'A' or 'B' microcode, load from tape, etc, etc

      --
      They sentenced me to twenty years of boredom
    8. Re:OH NOES!!! by CompMD · · Score: 1

      I'm just waiting for the day my 9406-500 that I use as a TV stand decides to roll itself over to my bed in the middle of the night and crush me, declaring, "VARY OFF HUMAN" in a cold, condescending voice.

    9. Re:OH NOES!!! by zzyzyx · · Score: 1

      Except it looks carefully crafted to be unpronounceable by a human pharynx ... This was planned all along !

    10. Re:OH NOES!!! by dwywit · · Score: 1
      I was going to use mine as a freezer - I bought it from ex-employer for AUD$10. Yep, a working AS/400 for ten bucks. Couldn't use it, though - the OS was licenced to the original owner and it wasn't transferable - and I didn't have enough power to make it go.

      So I gutted it (hung the CPU card on the wall as a memento), and I was going to line it with insulation and install a compressor/evaporator/condenser, but I never got around to it.

      Now it stores home brew.

      --
      They sentenced me to twenty years of boredom
  4. Oh, yes. by AltGrendel · · Score: 4, Funny

    And also to find the question that "42" answers.
    There are many theories as to what this question might be, and now IBM is building a system that will solve this issue once and for all.

    --
    The simple truth is that interstellar distances will not fit into the human imagination

    - Douglas Adams

    1. Re:Oh, yes. by MightyYar · · Score: 4, Funny

      No, no. It's being used to calculate the new national debt.

      --
      W..w..W - Willy Waterloo washes Warren Wiggins who is washing Waldo Woo.
    2. Re:Oh, yes. by ukbazza · · Score: 2, Funny

      No, no. It's being used to calculate the new national debt.

      They're going to need a bigger computer.

    3. Re:Oh, yes. by morgan_greywolf · · Score: 1

      And programmers with a firm grasp of arbitrary-precision arithmetic.

    4. Re:Oh, yes. by Viking+Coder · · Score: 1

      No, it can process over 3 tax returns per day.

      --
      Education is the silver bullet.
    5. Re:Oh, yes. by Anonymous Coward · · Score: 0

      I believe i have your question:
      41 + 1

      Thank me later.

    6. Re:Oh, yes. by Anonymous Coward · · Score: 1, Funny

      That's just your theory.

    7. Re:Oh, yes. by AmiMoJo · · Score: 1

      I bet it's cracking AES encryption.

      Government wants to crack encrypted files of enemies. Realises that it can build a computer capable of doing it for some billions of dollars. Finds excuse to build such a machine.

      Although brute-forcing the entire AES keyspace is still infeasible, brute forcing every possible password consisting of 30 typeable characters long certainly isn't.

      --
      const int one = 65536; (Silvermoon, Texture.cs)
      SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
  5. 2012? by DodgeRules · · Score: 1, Offtopic

    Is this the real reason the world ends in 2012?

    1. Re:2012? by Hal_Porter · · Score: 5, Funny

      A group of computer scientists build the world's most powerful computer. Let us call it "HyperThought." HyperThought is massively parallel, it contains neural networks, it has teraflop speed., etc. The computer scientists give HyperThought a shakedown run. It easily computes Pi to 10000 places, and factors a 100 digit number. The scientists try find a difficult question that may stump it. Finally, one scientist exclaims: "I know!" "HyperThought," she asks "is there a God?" "There is now," replies the computer.

      --
      echo -e 'global _start\n _start:\n mov eax, 2\n int 80h\n jmp _start' > a.asm; nasm a.asm -f elf; ld a.o -o a;
    2. Re:2012? by Anonymous Coward · · Score: 0

      If I were registered, this would definitely deserve mod points.

    3. Re:2012? by tingeber · · Score: 1

      Fredric Brown, "Answer", 1954:
      http://www.roma1.infn.it/~anzel/answer.html

      --
      oh my god... it's full of stars!
  6. End of the world in 2012 by lukaszg · · Score: 2, Informative

    I've heard about predictions of the end of the World in 2012, now I know the answaer - this machine will become a Singularity.

    1. Re:End of the world in 2012 by MozeeToby · · Score: 1

      In all seriousness, how much processing power would it take to run a program that designs newer and better processors? I would think that 20 petaflops and a good algorithm would be able to produce a processor that is an improvement over the current generation. Then again, I know next to nothing about processor design, so I could be totally wrong.

    2. Re:End of the world in 2012 by nicodoggie · · Score: 1

      That or the LHC makes one first.

    3. Re:End of the world in 2012 by Anonymous Coward · · Score: 0

      In all seriousness, how much processing power would it take to run a program that designs newer and better processors? I would think that 20 petaflops and a good algorithm would be able to produce a processor that is an improvement over the current generation. Then again, I know next to nothing about processor design, so I could be totally wrong.

      If we could make a computer capable of designing a better one, we'd already have the computer we wanted.

    4. Re:End of the world in 2012 by Anonymous Coward · · Score: 0

      WHOOSH.

      We don't have computers that can do real-time raytracing at decent resolutions, yet we can appear to have that by prerendering it.

      This is possibly the silliest thing i have read on Slashdot in a long time.

    5. Re:End of the world in 2012 by kalirion · · Score: 1

      The problem is not the hardware, it's the software. Who's going to write the initial algorithm?

      Ok, given enough processing power you could do a genetic algorithm for processor design that actually provides useful solutions within a reasonable amount of time, but I have a feeling we're far from that point.

    6. Re:End of the world in 2012 by jvkjvk · · Score: 1

      You're right, we have the horsepower.

      Now, if only we could produce the software Y'know, the set of "good algorithms" to produce the layouts and the other set of "good algorithms" to test fitness and ... everything else necessary to automatically produce solutions.

      *That* seems to be the hard task at the moment. I don't design processors either, so I don't know what types of issues the current design software has but to me it seems that this is probably the hurdle they are facing on that front, not processor power.

    7. Re:End of the world in 2012 by harry666t · · Score: 2, Interesting

      Hm, it's all about getting the right fitness function, isn't it?

      The processor that would be more fit would: draw less power, compute stuff faster, be cheap to produce, etc. Then it could either have a compatible instruction set, or a new one; in case of a new one, it would have to be able to come up with a way of automatically translating stuff from the old instruction set, or targetting a compiler at it.

      The case with the new instruction sets sounds really, really interesting. I think the actual hardware design would have to be derived from some higher-level representation of the instruction set architecture. I wonder if such a high-level description would be enough to also automatically port a compiler. Hmmmm...

    8. Re:End of the world in 2012 by kalirion · · Score: 3, Informative

      Well, whatever the case, once such a computer is built, someone had better ask it whether or not entropy can be reversed.

    9. Re:End of the world in 2012 by Timoleon · · Score: 0

      Well, whatever the case, once such a computer is built, someone had better ask it whether or not entropy can be reversed.

      The Last Question?

    10. Re:End of the world in 2012 by Anonymous Coward · · Score: 0

      The Last Question?

      INSUFFICIENT DATA FOR MEANINGFUL

      ........ silly lameness filter doesn't know when it's not needed

    11. Re:End of the world in 2012 by Anonymous Coward · · Score: 0

      someone had better ask it whether or not entropy can be reversed

      INSUFFICIENT DATA FOR MEANINGFUL ANSWER

  7. Second Post! by Anonymous Coward · · Score: 0

    Damn! So close. If only I had less of a life...

  8. at least they admit its true purpose by khuber · · Score: 3, Funny

    "The system will also act as a giant weather cock,"

    1. Re:at least they admit its true purpose by v1 · · Score: 5, Funny

      Primarily to keep track of nuclear waste

      And this can't be done with say, Excel?

      --
      I work for the Department of Redundancy Department.
    2. Re:at least they admit its true purpose by vlm · · Score: 5, Insightful

      And this can't be done with say, Excel?

      Ahh, Excel... the first choice in corporate database management systems.

      How many other slashdotters work at fortune XXX firms where on paper some executive bean counter says "we use oracle" but on the ground all databases are done in Excel (along with a smattering of everything else?)

      It is a step up from three jobs ago, where at another fortune XXX the database management system of choice was what boiled down to an administrative assistant and Lotus's word processing solution. Yes we used plain english to request that Patti make changes instead of sql update statements. Also our sql select statements always began with "hey Patti, could you look up...". Any yes, all "ORDER BY" stanzas were in fact powered by swear words and performed by cut and paste.

      Sadly I am not making any of this up.

      --
      "Science flies us to the moon. Religion flies us into buildings." - Victor Stenger
    3. Re:at least they admit its true purpose by Anonymous Coward · · Score: 0

      And this can't be done with say, Excel?

      No it can't, because this machine doesn't have enough horsepower to run Vista.

    4. Re:at least they admit its true purpose by this+great+guy · · Score: 1

      And this can't be done with say, Excel?

      RTFA that's the point ! They are building this supercomputer to run Excel 2012.

    5. Re:at least they admit its true purpose by evilbessie · · Score: 1

      No be fair there are an awful lot of Access 2 DBs* out there, Excel being a spreadsheet and all NOT a database.

      *oh so useful and forwards compatible

    6. Re:at least they admit its true purpose by Anonymous Coward · · Score: 0

      yes everything operational is done with Oracle, SAP, etc. But everything strategic is done in Excel.

  9. don't smell right by Trailer+Trash · · Score: 3, Funny

    Each processor gets its own megabyte of memory? Are these a bunch of refurb pcs from the late 80's?

    1. Re:don't smell right by Viking+Coder · · Score: 1

      Yeah, something's not right here. I came to the same conclusion.

      Did someone tell the author that they had that much L1 memory, and they didn't understand the difference?

      --
      Education is the silver bullet.
    2. Re:don't smell right by belrick · · Score: 1

      1.6 million processors in 96 racks is 16000 processors per rack, or about 400 processors per U. To me that sounds like an evolution of the cell processor and 1 MB per cell sounds reasonable.

    3. Re:don't smell right by 77Punker · · Score: 1

      It's SIMD. It's a simple set of instructions running on a whole lot of little sets of data. This is the same thing video cards do and it is a great way to solve (some) problems.

    4. Re:don't smell right by jsiples · · Score: 2, Informative

      The ram was wrong, its actually 1.6PB instead of TB

      --
      http://siples.kicks-ass.net
      The World is my Oyster
    5. Re:don't smell right by Anonymous Coward · · Score: 0

      What? Like a PS/2 running OS/2?

    6. Re:don't smell right by prograde · · Score: 1

      1.6 million processor / 96 racks / 43 U/rack

      = 388 processors/U

      clearly, we're not in Kansas anymore.

  10. Crysis? by thebheffect · · Score: 1, Redundant

    No Crysis comments yet? None? Ok.... But does it run Crysis?

    1. Re:Crysis? by Joce640k · · Score: 0, Offtopic

      And what about Windows Vista. Will we finally have a machine which can run Vista at interactive rates?

      --
      No sig today...
    2. Re:Crysis? by Anonymous Coward · · Score: 0

      Can it run Windows 7?

    3. Re:Crysis? by SirCarloz · · Score: 1

      I thought it could but Windows 7 requires 2.6TB of ram to run the new windows live messenger.

      --
      Echo Sexy Nerd :)
    4. Re:Crysis? by eulernet · · Score: 1

      Imagine 1.6 million UAC popups opening when you run a single program...

    5. Re:Crysis? by vistapwns · · Score: 1

      Beats 1.6 million password prompts...

      --
      "...I think the Microsoft hatred is a disease." - Linus Torvalds
    6. Re:Crysis? by makapuf · · Score: 1

      imagine running "top" with 1.6 million CPU lines refreshed each 3 seconds (or taskmaganer or cat /proc/cpuinfo)

  11. One word: by Anonymous Coward · · Score: 0

    ECHELON
    Keep track of nuclear waste, my ass!

  12. Processors, not cores by TeknoHog · · Score: 4, Insightful

    Because, when you put two processors on a single piece of silicon, it magically becomes one "processor" with two "cores".

    --
    Escher was the first MC and Giger invented the HR department.
    1. Re:Processors, not cores by MightyYar · · Score: 2, Funny

      Yeah, but how many CPUs is that?

      --
      W..w..W - Willy Waterloo washes Warren Wiggins who is washing Waldo Woo.
    2. Re:Processors, not cores by CannonballHead · · Score: 1

      Mmm, doesn't the difference between a core and a processor have to do with how they are connected?

    3. Re:Processors, not cores by evilbessie · · Score: 1

      Yes, yes it does. The issues whilst similar for multi-socket and multi-core systems are different due to the single processor having links to system bus and main memory shared between the cores, where as these are separate links on different processors. So as nomenculature goes it is not that bad at all.

    4. Re:Processors, not cores by TeknoHog · · Score: 1

      The issues whilst similar for multi-socket and multi-core systems are different due to the single processor having links to system bus and main memory shared between the cores, where as these are separate links on different processors. So as nomenculature goes it is not that bad at all.

      I agree it's useful to know what kind of multiproc system you're dealing with. However, the tone of this summary (IMHO) is that multisocket is somehow inherently better than the same number of cores in multi-CPU dies.

      On the other hand, I've read a number of opinions stating exactly the opposite. It really depends on the application; faster interconnects between CPUs vs. dedicated cache and outerconnects.

      On a side note, most of the time I see the word 'core' is in a context completely ignorant of multiprocessing history. There actually was SMP before Intel invented its Core, there were challenges on how to utilize multiple CPUs concurrently, and they were solved to a great extent.

      --
      Escher was the first MC and Giger invented the HR department.
  13. And for the opposite of petaflops... by Anonymous Coward · · Score: 0

    ...see here.

  14. 1.6M Processors, but only 1.6 TB memory? by fotbr · · Score: 1

    I would have expected it to have a bit more memory with that many processors.

    1. Re:1.6M Processors, but only 1.6 TB memory? by Professeur+Shadoko · · Score: 5, Informative

      My bet is that this is a typo.
      1.6 PB seems more reasonable.

    2. Re:1.6M Processors, but only 1.6 TB memory? by jbeaupre · · Score: 1

      Heck, I'd expect the total cache to be bigger than 1.6 TB

      --
      The world is made by those who show up for the job.
    3. Re:1.6M Processors, but only 1.6 TB memory? by rubycodez · · Score: 5, Informative
    4. Re:1.6M Processors, but only 1.6 TB memory? by countvlad · · Score: 1

      Hell, I'd like to know how 1.6 million processors can fit in just 96 racks. That's more than 16 thousand processors per rack. Either these aren't standard microprocessors or someone got their numbers a couple of orders of magnitude off.

      Even if you could get 16000 modern server processors into a rack, the power density alone would probably cause it to burst into flames (>1.6 MW)!

    5. Re:1.6M Processors, but only 1.6 TB memory? by thpr · · Score: 1

      I wouldn't. They explicitly choose algorithms that save memory and reduce communications requirements, even if that wastes CPU time. That is rational based on overall system power usage. It is also one of the design characteristics of IBM's BlueGene system, and I would expect that to hold for this system as well.

    6. Re:1.6M Processors, but only 1.6 TB memory? by Anonymous Coward · · Score: 0

      You have 2 TB of RAM in your desktop?

      Colour me impressed.

    7. Re:1.6M Processors, but only 1.6 TB memory? by thpr · · Score: 1
      Contrary to the original post, these are probably processor cores, not processors. The article doesn't clarify which is being talked about.

      Consider that IBM's original BlueGene was 1M processor cores in 96 racks (and designed in 1997) and 1.6M cores seems in the range of what I'd expect.

    8. Re:1.6M Processors, but only 1.6 TB memory? by Madball · · Score: 5, Informative

      Another reference article: http://www.eetimes.com/news/design/showArticle.jhtml?articleID=213000489 Mentions "up to" 4,096 processors per rack. So, at maximum, this would be 393,216 processors. Perhaps they are quad cores and someone took the liberty of multiplying the 393,216x4=1.6M (rounded). A more reasonable assumption may be 100,000 quad-core CPUs (400,000 cores). That would make the summarization of by only 16 times, lol.

    9. Re:1.6M Processors, but only 1.6 TB memory? by Taliesin · · Score: 1

      This EE Times article confirms that it's 1.6 PB. http://www.eetimes.com/news/design/showArticle.jhtml?articleID=213000489

      It also says they'll be 16 cores per processor!

    10. Re:1.6M Processors, but only 1.6 TB memory? by Anonymous Coward · · Score: 0

      I liked the original number of processors better. One for each Linux Distro. Now they're SOL.

    11. Re:1.6M Processors, but only 1.6 TB memory? by Anonymous Coward · · Score: 0

      "BlueGene/P uses a modified PowerPC 450 processor running at 850 MHz with four cores per chip and as many as 4,096 processors in a rack. The Sequoia system will use 45nm processors with as many as 16 cores per chip running at a significantly faster data rate.

      Both BlueGene/P and Sequoia consist of clusters built up from 96 racks of systems. Sequoia will have 1.6 petabytes of memory feeding its 1.6 million cores, but many details of its design have not yet been disclosed."

    12. Re:1.6M Processors, but only 1.6 TB memory? by Madball · · Score: 1

      "BlueGene/P uses a modified PowerPC 450 processor running at 850 MHz with four cores per chip and as many as 4,096 processors in a rack. The Sequoia system will use 45nm processors with as many as 16 cores per chip running at a significantly faster data rate.

      Both BlueGene/P and Sequoia consist of clusters built up from 96 racks of systems. Sequoia will have 1.6 petabytes of memory feeding its 1.6 million cores, but many details of its design have not yet been disclosed."

      There we go. It is 100,000 processors, with 16 cores each (yes, a core is a processor, but since the summary went out of its way to make this distinction, we should continue to do so for a fair comparison). Summary is wrong (big surprise there).

    13. Re:1.6M Processors, but only 1.6 TB memory? by hydromike2 · · Score: 1

      200,000 octcacore xeons would not be unreasonable, unless they are counting cores like in the newer nvidia cards in which case it gets much easier to imagine 4096 cores per rack

    14. Re:1.6M Processors, but only 1.6 TB memory? by BikeHelmet · · Score: 1

      You bring up an interesting point.

      A while ago there was a company that produced energy-efficient supercomputers. The motherboards they used were absolutely packed with low-power low-speed fanless CPUs.

      I don't think they survived - not enough demand - most companies prefer faster, fewer cores.

      I believe their smallest system had something like 96 CPUs, and biggest about 3500. Power consumption was around 1 watt for each CPU, plus whatever the RAM uses. That means in theory you could run a 3500 CPU beast in your own home - at the loss of several rooms.

      Good luck coding for it. :D

  15. Aluminum foil hat. by Samschnooks · · Score: 1, Offtopic

    Besides managing stockpiles of rusty old nukes, Sequoia will also be used for research into astronomy, energy, the human genome and climate change, according to IBM.

    It could also be used to search for "suspicious behaviour" by searching Government databases, Credit card companies' databases, credit bureau databases, Choicepoint's, telecommunication companies' databases, airlines, and any other firm that the Government bullies into giving access. Own a gun, buy some grow lights for your reef tank, and fertilizer for your spinach fields and that'lll be a searching and detainment because of the "War on Drugs", "War on Terror", and "War on [insert here]".

    Will we be allowed any sort of control over our own information? I think not!

    1. Re:Aluminum foil hat. by wild_quinine · · Score: 2, Insightful

      It could also be used to search for "suspicious behaviour" by searching Government databases, Credit card companies' databases, credit bureau databases, Choicepoint's, telecommunication companies' databases, airlines, and any other firm that the Government bullies into giving access.

      Well, that's not as paranoid as you might think. The case against is quite simply the publicity that's been given to this behemoth of a machine, so I really don't think it's too likely in this particular case.

      However this is EXACTLY how you go about putting together a machine for intelligence purposes. The key to running an intelligence service is deniability at as many levels as possible, and keeping anyone from seeing the big picture.

      So you comission some huge piece of hardware, with a benign-but-complex sounding (usually simulation) function.

      Then you get the low leve software put together for the platform. If that can be done in modular fashion, so much the better. You don't mix the platform with the real world data during design.

      At the final stages, and presumably in house, you can write your overlaying interface (which intelligence employees will use), and only then is the pure function of the suite necessarily made apparent.

      Of course there are lots of people in the design process that have a notion of how things are being put together, what they will interface with, etc, who can take a stab at the function. And sure, everyone signs an NDA just because. However, since nobody sees all of it, and the big picture is never confirmed outside of a very small number of people, and nobody is going to break an NDA to talk about part of something that MAY have some function... you essentially reduce the risk of leaking your cababilities, system spec, and intention by a not insignificant amount.

    2. Re:Aluminum foil hat. by Spatial · · Score: 1

      Yeah. Look at what Sweden uses their magnificent supercomputer for. Hopefully you guys won't go (further) down that road.

    3. Re:Aluminum foil hat. by Rhys · · Score: 4, Insightful

      You don't need 20 petaflops to do that, you need a few tens of teraflops and a really really huge memory and really really fast IO. You'd do much better with some of the 1/4TB memory systems from Sun or IBM + spending a huge pile of money on SSDs than a real supercomputer.

      The cost of the IO interconnect is a huge chunk of cash to sink into a supercomputer that you just don't need for that sort of tin foil hat application.

      --
      Slashdot Patriotism: We Support our Dupes!
    4. Re:Aluminum foil hat. by bockelboy · · Score: 2, Insightful

      I'm sorry, I don't believe it.

      I think using a BlueGene for run-of-the-mill data processing would be a horrible waste of money. There's simply no need for things like a parallel filesystem or PB of RAM or low-latency interconnects. You want to "scale out" for distributed processing like you're talking about, not "scale up".

      No, I'd bet intelligence gathering is done on Google-like processor farms.

    5. Re:Aluminum foil hat. by Creepy+Crawler · · Score: 0, Troll

      Yeah. Look at Pirate bay.

      XD

      --
    6. Re:Aluminum foil hat. by pipatron · · Score: 1

      This computer would however be really good at brute-forcing crypto keys...

      --
      c++; /* this makes c bigger but returns the old value */
    7. Re:Aluminum foil hat. by khallow · · Score: 1

      Frankly, that's what worms and zombie networks on the internet are for when they aren't sending spam.

    8. Re:Aluminum foil hat. by Anonymous Coward · · Score: 0

      Nice tinfoil hat, but given the stated purpose of nuclear research, I bet this machine will be on the "Secure" network. That means the network it is on has no physical connection to the outer world whatsoever (to eliminate any possibility of hacking). It and the terminals which can access it will be completely isolated. Therefore, it would be quite hard for it to listen in on anything other than itself and the other supercomputers on the "secure" side.

      ~Former LLNL employee

    9. Re:Aluminum foil hat. by Spatial · · Score: 1

      Nah... That's a sort of distributed computing endeavour rather than a single supercomputer...

    10. Re:Aluminum foil hat. by evilbessie · · Score: 1

      That's way too public for IBM to release any infomrmation. Much easier for the government to just pay stacks o'cash to them to keep schtum about it.

    11. Re:Aluminum foil hat. by bnenning · · Score: 1

      This computer would however be really good at brute-forcing crypto keys...

      Not really, 2^N gets big fast. The sun won't output enough energy over its entire lifetime to allow a maximally efficient computer to even count from 0 to 2^256, let alone try to brute-force a 256-bit key. (From Applied Cryptography which I don't have in front of me).

      --
      How to solve most of our problems: 1.Lots of nuclear plants. 2.Cure aging.
    12. Re:Aluminum foil hat. by religious+freak · · Score: 1

      Agreed, this is a processing power beast, not an IO beast (at least from the details released). Tinfoil hat theory would suggest this is for "realtime" analysis of ... [insert favorite conspricy here].

      I'm sold. By my estimation, this is obviously not for "weather monitoring" and "climate change".

      --
      If you can read this... 01110101 01110010 00100000 01100001 00100000 01100111 01100101 01100101 01101011
    13. Re:Aluminum foil hat. by Gamma747 · · Score: 1

      Maybe they won't be able to generate 2^256 encryption keys, but they could try to factorize someone's public key. They could also try encrypting every possible string of a certain length with someone's public key and see which one produces the same ciphertext, generating every possible GPG signature and verifying them until they have a match, or brute-forcing hash algorithms. It's also possible that they've found a way to break encryption without generating 2^256 keys; maybe some encryption products have a flaw that drastically reduces the keyspace.

    14. Re:Aluminum foil hat. by bnenning · · Score: 1

      Maybe they won't be able to generate 2^256 encryption keys, but they could try to factorize someone's public key.

      Factoring is also (believed to be) exponential, so increasing the key length to a still-manageable size would be sufficient.

      They could also try encrypting every possible string of a certain length

      Only up to a few hundred bits.

      It's also possible that they've found a way to break encryption without generating 2^256 keys; maybe some encryption products have a flaw that drastically reduces the keyspace.

      That's quite likely, although then it wouldn't really count as "brute force".

      --
      How to solve most of our problems: 1.Lots of nuclear plants. 2.Cure aging.
    15. Re:Aluminum foil hat. by Anonymous Coward · · Score: 0

      I am not looking forward to 'War on [insert here]'

  16. MUST NOT BE USED FOR MOON RESEARCH by UbuntuLinux · · Score: 0

    Since it has become evident that the moon is haunted, hence the presence of such a computer on the moon would be a disaster for all mankind; ghosts do not understand technology, even those that haunt the moon, and if one where to find this computer, it could type in the wrong commands and cause terrible things to happen.

  17. flops not flop by iYk6 · · Score: 5, Funny

    flops = floating point operation per second
    flop = Gigli

    The article got it mostly right. It mentioned 500-teraflop once, but every other time it spelled flops correctly. Slashdot, on the other hand, fucked up the title, despite the fact that it pretty much just copied it from the article (poorly).

    1. Re:flops not flop by elrous0 · · Score: 1

      Don't you blaspheme Ben Affleck in this home, young man!

      --
      SJW: Someone who has run out of real oppression, and has to fake it.
    2. Re:flops not flop by Chris+Mattern · · Score: 1

      So what does it mean when they talk about them Gigliflops?

    3. Re:flops not flop by Chris+Mattern · · Score: 1

      No, genius. His point is that "flops" is singular, with the s standing for second rather than forming a plural. His point is that you say 2 petabyte hard drive and 500 teraflops machine because those are the singular forms.

    4. Re:flops not flop by Anonymous Coward · · Score: 0

      That's short for epic fail.

  18. so they can play raytraced quake mods by sgt+scrub · · Score: 2, Funny

    Hopefully the government uses this magnificent tool wisely when it gets it in 2012.

    Sounds like they are going to port the quake mods to the raytrace q4 engine.

    --
    Having to work for a living is the root of all evil.
  19. Nice by SirCarloz · · Score: 0, Troll

    Arent these the specs require by Windows 7 for home use ?

    --
    Echo Sexy Nerd :)
    1. Re:Nice by Anonymous Coward · · Score: 0

      Fail

  20. Yawn. by Anonymous Coward · · Score: 0

    It still does not reach the power of the worldwide google cluster.

  21. Why always nuclear simulation? by TheDarkMaster · · Score: 1

    Why always nuclear explosions simulation is the primary use for this type of computer? They do not have better uses like climate simulation, folding proteins or play crysis at full?

    --
    Religion: The greatest weapon of mass destruction of all time
    1. Re:Why always nuclear simulation? by QuantumRiff · · Score: 1

      Would you rather them set off nukes to study these things? The reason is that it takes a crapload of calculations to map out every reaction between molecules in an area measured in square miles. (and because of a test ban treaty we signed, we can't set off 'real' nukes to test anymore, so we have to simulate it')

      --

      What are we going to do tonight Brain?
    2. Re:Why always nuclear simulation? by Anonymous Coward · · Score: 0

      Because simulated nuclear explosions are less harmful than actual nuclear explosions. Besides according to TFA, that's not the only thing it will be used for.

    3. Re:Why always nuclear simulation? by Troy+Baer · · Score: 1

      Why always nuclear explosions simulation is the primary use for this type of computer?

      Uh, because it's paid for out of the NNSA budget?

      --
      "My life's work has been to prompt others... and be forgotten." --Cyrano de Bergerac
    4. Re:Why always nuclear simulation? by joib · · Score: 1

      Uh, do you know how many molecules there are in "square miles"? Do you know how many atoms you can calculate reactions between using both state of the art supercomputers and quantum chemistry tools? Do you know the scaling behavior of quantum chemistry methods? I mean, even with this new supercomputer, your estimate is off by ridiculously many orders of magnitude. And that being said, nuke simulation has little to do with quantum chemistry anyways.

    5. Re:Why always nuclear simulation? by Muad'Dave · · Score: 1

      Why always nuclear explosions simulation is the primary use for this type of computer?

      Would you rather they test nuclear explosions for real?

      --
      Tiller's Rule: Never use a word in written form that you've only heard and never read. You will end up looking foolish.
    6. Re:Why always nuclear simulation? by cbiltcliffe · · Score: 2, Insightful

      Because funding for military expenditures is much easier to obtain than funding for climate research.

      --
      "City hall" in German is "Rathaus" Kinda explains a few things......
    7. Re:Why always nuclear simulation? by ConanG · · Score: 1

      The real question is why are we still designing new nukes? Do they go obsolete after a few years? What are they tweaking the designs for? Better yield? We can send the planet into a nuclear winter already with what we've got.

    8. Re:Why always nuclear simulation? by Cjstone · · Score: 1

      IIRC, current nuke development is going towards shrinking the devices and minimizing fallout.

    9. Re:Why always nuclear simulation? by Richard_at_work · · Score: 4, Informative

      Yes, nuclear weapons have a shelf life due to the components included in them - explosives, chemicals etc.

      New designs are used to maximise yield per mass, enabling you to throw a smaller warhead at a target, which means less chance of interception. It also means a smaller package to maintain, and cheaper to build, along with more warheads per unit of material.

    10. Re:Why always nuclear simulation? by Bigjeff5 · · Score: 1

      And that being said, nuke simulation has little to do with quantum chemistry anyways.

      So why did you bring it up? The parent didn't, I don't get what you are saying when you ask a question, relate it to the parent's post, and then say it is irrelavent. You might as well have asked him if he knows how tight the car you drive corners, if you are going to say it is irrelavent anyway.

      And, do you realize how much processing power 20 petaflops is? That's insane, I'm having a hard time wrapping my head around it. That is well into the territory of the number of molecules in a small object. There are roughly 6.022x10^23 entities per mole, and at 2x10^16 calculations per second, simulating what happens in the first few instances of a nuclear explosion in a realistic time frame becomes more feasable. It would still be slow as all getout, and the most you'd ever want to calculate is less than a second for the individual atoms in a small object, but it's in the realm of possibility. You're right about square miles, but getting into that range there isn't really a need for tracking individual molecules, so that point seems moot to me.

      Also, they so far have not needed to calculate what a nuclear bomb does for each atom (obviously, since it has been nigh impossible), and they probably won't ever need to really. You can study waves and energy effects in great detail, and simulate them accurately, without needing to know where each and every atom goes. This will simply let them be more precise and accurate, as well as speedy.

      --
      Security is mostly a superstition... Avoiding danger is no safer in the long run than outright exposure. - Helen Keller
    11. Re:Why always nuclear simulation? by joib · · Score: 4, Informative
      So why did you bring it up

      The parent said that the computer will be used for "mapping every reaction" between molecules. Presumably, since reactions tend to require quantum mechanical descriptions, I guessed the parent meant that the new computer would allow doing such calculations for all reactions in a rather large area.

      I don't get what you are saying when you ask a question, relate it to the parent's post, and then say it is irrelavent.

      Just a gedanken experiment to amuse myself, while noting that it actually has nothing do with simulating nuclear weapons. Don't get too worked up about it.

      And, do you realize how much processing power 20 petaflops is?

      Yes, it's about 2 orders of magnitude more than the supercomputer I'm using at the moment. A lot for sure, but still limited to very small system sizes for quantum mechanical calculations. At the moment, even the best methods in practice scale as N^3 or so. With my current 100 TFlops I might do a DFT calculation with O(10000) atoms or so. Two orders of magnitude more CPU power with N^3 scaling gives me roughly a factor of 5 more atoms. 50000 atoms fit into a box of roughly 10x10x10 nm (depending on the material etc., of course). Still a way to go until I'm able to do "square miles"..

      If you want to go into classical molecular dynamics, then you're obviously in much better shape. With the current supercomputer that's maybe around 1E9 atoms, and since MD scales linearly, with two orders of magnitude more flops it means around 1E11 atoms. Now these fit into a box on the order of 1 um**3. Again, still quite a way to go to square miles..

      In conclusion, atoms are really really tiny, and in 3 dimensions you can pack a lot of them into a very tiny volume.

      Also, they so far have not needed to calculate what a nuclear bomb does for each atom (obviously, since it has been nigh impossible), and they probably won't ever need to really. You can study waves and energy effects in great detail, and simulate them accurately, without needing to know where each and every atom goes. This will simply let them be more precise and accurate, as well as speedy.

      Yes, that was sort of implied in my previous post. The US nuke labs have been at the forefront in research on numerical methods in topics such as shock propagation (PPM and methods like that) and really really large FEM simulations. Obviously, the actual nuclear reactions are taken into account probabilistically rather than the full quantum mechanical treatment (as my above monologue shows, such a treatment for the primary is far beyond any computer in sight). AFAIK they use Monte Carlo neutron diffusion rather than the classical multigroup diffusion methods that AFAIK are still largely used for civilian reactor design. That being said, I'm sure they are doing a lot of atomic and quantum level simulations as well for small model systems designed to e.g. extract parameters for continuum simulations and such.

    12. Re:Why always nuclear simulation? by Anonymous Coward · · Score: 0

      One useful goal would be to make them cleaner, engineering them so that the reaction is more complete or produces different byproducts with half lives of hours or days rather than years. We might be able to produce the same destruction with fewer or at least smaller, less maintenance intensive devices. Tritium is possibly the most expensive material by weight in the world, so anything that could be done to reduce how much we need in the doomsday devices would be quite useful. Also, any work on the related launchers would be useful for space stuff as well.

    13. Re:Why always nuclear simulation? by BigAssRat · · Score: 1

      Because noone can predict what the climate is going to do. Why bother running simulations of the climate on this thing? So we can run bad attempts at prediction faster?

    14. Re:Why always nuclear simulation? by Anonymous Coward · · Score: 0

      >"...nuclear weapons have a shelf life due to the components included in them"
      Actually they have a shelf life because the half-life of tritium is 12.3 years.

    15. Re:Why always nuclear simulation? by TuballoyThunder · · Score: 1
      There are two basic reasons:
      • If you replace a component or change the nuclear explosives package, you need to determine the effect on the performance
      • If you notice a problem with a weapon during routine surveillance, you need to determine if it is significant
  22. cores not cpu's by Anonymous Coward · · Score: 0

    sorry eldavojohn, you got it completely wrong in your summary. From eweak article:

    "98,304 compute nodes and 1.6 million processing cores."

    1000 CPU's per rack makes alot more sense.

  23. "said" missiles? by Anonymous Coward · · Score: 0

    There are no missiles mentioned anywhere. Does anybody check these simple blurbs before submitting them?

    1. Re:"said" missiles? by Anonymous Coward · · Score: 0

      Hey, it's enough of a struggle just for the "editors" to figure out how to click the "Accept" button.

      I just had an idea. The "editors" are border-line retarded. George W. Bush is border-line retarded. Slashdot should hire GWB! Pudge would finally have someone to hang around with and the "editorial" quality would slightly improve. Everybody wins!

  24. Skynet anyone? by Anonymous Coward · · Score: 0

    With almost 1 billion transistors per microprocessor these days, this super computer with over 1 million microprocessors is exceeding the 100 trillion synaptic connections in an average human brain.

    1. Re:Skynet anyone? by khallow · · Score: 2, Interesting

      Interesting. There's still the matter of the software. Currently, these machines are running straightforward simulation software. I don't see weather prediction resulting in sentience any time soon. The machines will probably have to grow considerably before the more flexible software subsystems like say the load-balancing code, achieves sentience, and destroys us all.

    2. Re:Skynet anyone? by H3g3m0n · · Score: 1

      The human brain isn't actually that complex to implement since a lot of it is self generated, IBMs Blue Brain project have already simulated a mouses neo cortex (the part of the brain that makes conscious thought). They have been claiming to scale it up to a human one by 2010 for a while now...

      --
      cat /dev/urandom > .sig
    3. Re:Skynet anyone? by khallow · · Score: 1

      Still sounds like a big stretch. Having said that, I think we have enough hardware for sentient computers, we just don't have the software.

  25. Fixed if for ya by Metabolife · · Score: 2, Funny

    "...allowing forecasters to create local weather "events" less than one kilometer across, compared with 10 kilometers today and at speeds up to 40 times faster than current systems."

  26. THIS IS ABSOLUTELY NECESSARY TO PROTECT US! by Anonymous Coward · · Score: 0

    How else will the government keep track of the prisoners housed in their interment camps for the new menace directly threatening not only America, but the entire human population! There are people amongst us, who are an imminent threat to the stability and safety of everything we hold dear. If it weren't for the efforts of my boss, the new junior Senator from New York, we would be facing a disaster which you can't even begin to comprehend! I've seen the reports of what these people can do, and their numbers are growing every day. Without this computer, and its ability to analyze massive amounts of data and scanning millions of DNA records, it will be impossible for us to track them down and detain them before they rise up and destroy everything which we hold dear.

    1. Re:THIS IS ABSOLUTELY NECESSARY TO PROTECT US! by jamstar7 · · Score: 1

      Right, that's the classified new menace, the one they're not allowed to talk about but need a few trillion bucks to combat.

      --
      Understanding the scope of the problem is the first step on the path to true panic.
  27. Who is building it? by Anonymous Coward · · Score: 1

    China? IBM is laying off all the Americans and Europeans.

    1. Re:Who is building it? by religious+freak · · Score: 1

      No, they're only providing the hardware with the embedded viruses and spyware to us - free of charge!

      --
      If you can read this... 01110101 01110010 00100000 01100001 00100000 01100111 01100101 01100101 01101011
  28. Unfortunate Typo by IRIGHTI · · Score: 1

    "The system will also act as a giant weather cock, allowing forecasters to predict local weather "events" less than one kilometer across,..." Hehe.

    1. Re:Unfortunate Typo by Andy_R · · Score: 1

      It's not a typo, a weather cock is just another name for a weather vane, it refers to the traditional shape of weather vanes

      --
      A pizza of radius z and thickness a has a volume of pi z z a
    2. Re:Unfortunate Typo by Areyoukiddingme · · Score: 1

      It's not a typo. A weather cock is a wind vane in the shape of a rooster. They were exceedingly common on barn roofs for decades (if not centuries) and are still fairly common today. Wikipedia image. They tend to be inaccurate in low windspeeds because they're relatively heavy and their pivots are often poorly maintained.

  29. So let's see.... by cbiltcliffe · · Score: 5, Insightful

    - IBM is building a computer that will be functional in about 3.5 years.

    - The power of this computer, in 3.5 years, will outshine every other supercomputer currently running today.

    I should hope so! What's the point of taking 3.5 years to build the thing, if it's going to be 3.5 years out of date by the time they build it?

    Heck, in 3.5 years, your desktop computer will be 4 times more powerful than anything currently running today, too.

    Duuh.

    --
    "City hall" in German is "Rathaus" Kinda explains a few things......
    1. Re:So let's see.... by SirCarloz · · Score: 1

      Very true, AMD and intel cpus are already running at 6ghz with proper cooling , that'll be running on home computers no later than this year.

      --
      Echo Sexy Nerd :)
    2. Re:So let's see.... by m0i · · Score: 1

      If you have a look at http://top500.org/lists/2008/11/performance_development it takes more than 6 years to get 10 times actual performance (quicker than Moore's law, hrm). Given that the actual top is at 1PF, going to 20 in 3.5 years is quite an achievement.

      --
      have you been defaced today?
    3. Re:So let's see.... by CannonballHead · · Score: 1

      Heck, in 3.5 years, your desktop computer will be 4 times more powerful than anything currently running today, too.

      For being so picky about the terms in the article, you are quite lax with your own. I seriously doubt my desktop, in 3.5 years, will be able to do ~ 6 petaflops. :) (4x more powerful than "anything" currently running today)

      Furthermore, 20 vs. ~1.5 petaflops is a goodly sized jump for 3ish years, isn't it? Computer speed growth has seemed to be slowing lately, with an emphasis being on multiple cores, not faster clock speeds like it was 10 years ago. So being able to get 20x the power of the current super computer in 3.5 years seems to be pretty good.

    4. Re:So let's see.... by cbiltcliffe · · Score: 1

      It's obvious that I was referring to desktop computers with the "anything running today" wording.

      If you're going to be that intentionally disingenious, why don't you also say that I claimed desktop computers were going to have 4000+ horsepower, since there are industrial earth moving equipment engines that currently put out over 1000. ....wait a minute.....

      --
      "City hall" in German is "Rathaus" Kinda explains a few things......
    5. Re:So let's see.... by Anonymous Coward · · Score: 0

      In the actual article it said it would outshine everything on the top 500 list combined.

  30. Because we can accurately model by Shivetya · · Score: 2, Insightful

    nuclear explosions whereas climate simulations don't have all the variables.

    Actually I think they model the effects on decay in current nuclear weapons. Besides its not something I want them to physically test.

    --
    * Winners compare their achievements to their goals, losers compare theirs to that of others.
  31. 20 PFlops? Do the math by Anonymous Coward · · Score: 0

    Do the math

    1.6 million core * 850 MHZ * 4 flops/cycle is only 5.44 PetaFlops , not 20 PetaFlops. How/why does the article say 20 PetaFlop, what else is going on?

    1. Re:20 PFlops? Do the math by TheKidWho · · Score: 1

      Quad Cores?

    2. Re:20 PFlops? Do the math by DegreeOfFreedom · · Score: 1

      You seem to be assuming that IBM will use the PowerPC 450 chips that BG/P uses. Sequoia will instead use some not-quite-yet-disclosed 45 nm part with a much higher clock rate.

  32. Why does the government need it? by scorp1us · · Score: 1

    Why can't we let private industry own the computer and the government just purchase time on it? I for one would love to have CGI movies rendered in better-than-real time. This way, us the taxpayers don't have to pay for idle time.

    Also, I can design a database using SQLite with a web front end for keeping track of uranium or anything else for that matter. As long as it is not measured in individual atoms, it'll run fine on my spare 2.4 Single core celeron. There is no need to update the database 100M times a second.

    --
    Slashdot's rate-of-post filter: Preventing you from posting too many great ideas at once.
    1. Re:Why does the government need it? by rickb928 · · Score: 1

      You could probably do that with an old copy of Filemaker Pro and a eeePC.

      Sheesh. It's inventory management, not rocket science.

      --
      deleting the extra space after periods so i can stay relevant, yeah.
    2. Re:Why does the government need it? by religious+freak · · Score: 1

      Outside of maintenance periods, I would be highly surprised if there were any downtime on this thing. I'd imagine it will be booked out years in advance, just like most other supercomputers - and "super" telescopes.

      How much downtime does blue gene have? Hubble?

      --
      If you can read this... 01110101 01110010 00100000 01100001 00100000 01100111 01100101 01100101 01101011
  33. They've cracked RSA! by Arancaytar · · Score: 1

    Dan Brown told me so.

  34. typing in top - 1 would be funny by Ruede · · Score: 1

    typing in top -> 1 would be funny

  35. But.. by ZERO1ZERO · · Score: 1

    ...will it run Vista?

    1. Re:But.. by TheCybernator · · Score: 1

      2012 WILL be the year of Linux.

  36. Simpler solution by Anonymous Coward · · Score: 0

    Nice supercomputer and all. But drugs and a 5$ wrench would've worked better.

    1. Re:Simpler solution by Anonymous Coward · · Score: 0

      That's a lot of wrenches. Drugs, the CIA got those.

      Imagine a Beowulf cluster of $5 wrenches...

      And the humilty of realizing you can't run Linux on a $5 wrench. Not enough hexaflops.

      And the sheer multitasking necessary to wield all those wrenches. Oh my.

  37. Yah, right.... by Beat+The+Odds · · Score: 1

    Hopefully the government uses this magnificent tool wisely...

    Just like they do everything else...

  38. MTBF by vlm · · Score: 3, Interesting

    So the real question in an immense cluster like this, is whats the MTBF?

    Simon claims that the Eniac MTBF was 8 hours, although I've seen all kinds of claims on the web from minutes to days.

    http://zzsimonb.blogspot.com/2006/06/mtbf-mean-time-between-failure.html

    I would guess this beast will never be 100% operational at any moment of its existence.

    I'm guessing the "cool" part of this won't be the bottomless pile of hardware in one room, but how they maintain this beast. Just working around one of the million CPU fans burning out is no big deal, but how do you deal with a higher level problem like one of the hundreds of network switches failing, etc?

    --
    "Science flies us to the moon. Religion flies us into buildings." - Victor Stenger
    1. Re:MTBF by mmell · · Score: 3, Informative
      Higher than you're guessing. I've worked on BlueGene/L, BlueGene/S and was involved in some of the development on BlueGene/P. All of these systems have an incredibly agressive monitoring mechanism - voltages, temperatures, fan speeds, as well as half a dozen other hardware categories are monitored at the component level and the data stored in a database where it is analyzed to ensure that the system as a whole IS operational and stays that way.

      But thank you for pointing out that the architecture is inherently fault-tolerant. When submitting a job to massively parallel machines like this, one of the options presented is how many cores/how much memory for this job (as well as many other performance affecting options such as internal network topology). A bad core can be "skipped" in much the same way as a bad sector on a hard drive; except that in this case, it's possible to repair/replace the bad sector on the fly.

    2. Re:MTBF by joib · · Score: 1

      Yeah, but an MPI job can't recover from a failed node. Except for checkpointing, of course. So if you launch a job on all those 1.6e6 processors, they all better on average stay up at least long enough that you make some progress and write a checkpoint before one node crashes.

    3. Re:MTBF by mmell · · Score: 1

      Seldom (if ever) does a single job get all the processors. I think I only saw it when we were running LINPACK to see if we were hitting our numbers. That job runs for quite some time, and even when running on all 1000+ processors on our BlueGene I never once saw it fail due to hardware issues. In fact, in over a year I think I only saw one instance of a job having to be re-run due to hardware failures. I can't say my knowledge is exhaustive, but my personal experience is that the BlueGene supercomputer series has an excellent hardware availability record.

  39. Cheney's weather machine arrives too late?! by FatSean · · Score: 1

    Maybe Biden will resume the new tradition of VP as weather-manipulator.

    --
    Blar.
  40. wisely used? by furby076 · · Score: 1

    Hopefully the government uses this magnificent tool wisely when it gets it in 2012

    Yup, they will be the first people able to play Crysis.

    --

    I do not support "The Man". I also do not support your irrational stupidity
  41. Climate Change by Anonymous Coward · · Score: 0

    Investigate climate change....

    I bet this thing itself will cause climate change. Think it will get an energy star rating?

  42. FTFA by Anonymous Coward · · Score: 0

    "The system will also act as a giant weather cock"

    I for one welcome our giant weather cock overlords.

  43. Bullshit: +10, Informative by Anonymous Coward · · Score: 0

    "Hopefully the government uses this magnificent tool wisely when it gets it in 2012."

    The machine will ONLY be used for communications intercepts of EVERYTHING.

    I hope this helps your encryption program.

    Yours In Socialism,
    Kilgore Trout

  44. Hazards of too much processing power by Captain+Spam · · Score: 2

    Hopefully the government uses this magnificent tool wisely when it gets it in 2012.

    SCENE: The Pentagon, 2012

    Science Advisor: "President Whoever-You'll-Be, IBM has completed our 20 petaflop computer. It is awaiting your command."
    President Whoever-You'll-Be: "Thank you, Advisor. We can use it to compute the long-term effects of nuclear waste disposal, weather fronts, and... just... just how much processing power is in this?"
    SA: *deep sigh* "Over 1.6 million processors and a total of 1.6TB of RAM, sir."
    PWYB: "My GOD, Advisor. Do you know what that much power could do? It... it could..."
    SA: *another deep sigh* "It could, in theory, calculate the entire state of Wisconson to rubble. Or process the irrelevance of humanity down to a variance of 10^-24. Or, and this is what we were hoping not to worry you about, refactor..."
    PWYB: (interrupting) "Refactor the planet into a singularity, yes, I know. This is a grave situation. We can only hope this much processing power doesn't fall into the hands of someone with fixed-polarity Reed-Muller expressions for incompletely specified Boolean equations and a vendetta."
    SA: (long pause) "Shall I turn it on, sir?"
    PWYB: "Yes, turn it on. And may God have mercy on our souls if we never need an fsck..."

    --
    Demanding constant attention will only lead to attention.
  45. Re:Corrupt government by Anonymous Coward · · Score: 0

    U.S. government: Any amount of the taxpayer's money to research how to kill other people and destroy their property.

    And IBM can be relied upon to take any government's money to do exactly that.

  46. 3 years from now? by FLoWCTRL · · Score: 1

    "IBM reckons its 20-petaflops capable Sequoia system will outshine every single current system in the Top500 supercomputer rankings"

    So the computer will be ready in 2012, and it will outperform computers from 2009?

    These multi-year computer construction projects seem very problematic given the pace of change in technology. Memory changes, CPUs change, and the socket specs change — if it takes 3 years to build, it will be obsolete before it's ready. 2012 could be the year that ATI releases 10-petaflop GPUs, and they ship in iMacs.

  47. Can it balance the budget? by PackMan97 · · Score: 1

    The feds have been needed a computer that can balance a budget...think this monster is up to the task? Somehow I doubt it.

  48. They should just wait until 2012 by Anonymous Coward · · Score: 0

    And go pick one up off the Dell refurb site.

  49. The cost of Keeping Us Safe. by davonshire · · Score: 1

    When it is complete, it will be yet another amazing accomplishment of computer science, answering a call for a problem created by people who we elected or selected to help guard the citizens of the united states and the world.

    The problem of course isn't the focus of article. Nor are the resources this new answer will require made totally clear but at least it gives us a glimps of how much we're willing to spend to maintain the national vision of 'Safe'.

    The Hoover dam produces 2.8 Million kilowatts, this consumes about 0.11% of Hoover's capacity, now with the power conversion and distance to source you can expect that to pretty much quadruple in real energy terms. So let's say 0.44%, no mention of how much power the entire facility will need to operate it. (cooling, lighting, communications etc.) but I bet it's at least another 0.1% when properly tallied in.

    I think it's great that the government can spend this much money on a machine to help ensure the readiness of our amazing nuclear arsenal. But since this is an enhancment to another machine not long ago built to do the same thing. You have to wonder how much power, money, heat, carbon and whatever else we worry about will be used to build the next machine for the same purpose.

    The stockpile of our arsenal should be dwindling till we end up with a nice comfy number of weapons that can only wipe us out 2 or 3 times over.

    Does it make you wonder what could happen if this tool was used to help solve some of 'it's own problems?' the one that will ensure come 2012 there's enough power to power it?

    Let's hope the DOE has plans to use spare cpu power to find better materials for power production, conduction and storage. The quickest way to ensure prosperity is to have enough power to do all you need and some of your wants.

    There is not a single future world utopia that does not demonstrate a ubiquitous availibility of energy. Not Star Trek, Star Wars, Back to the Future II etc etc etc.

    While it's reasonable to consider lowering our energy needs in all things, the less we consume, the more, there should be for the future. It is no guarentee. Energy is one thing I do not think anyone can argue, life can exist without in some form or another.

    Fusion I prophetically believe is a blind alley that will never produce a sustainable energy credit. At least not in the way it's described currently.

    It's not just a question of being able to generate power either, we have to have cheap, economical and plentiful ways to convey the power and store it. We've done great things to increase the efficiency of the things that use the power we currently produce. But no matter how much better we make them eventually it will not outpace our current propensity to consume the power we make.

    Turn the problem on it's head and see if any change falls out.

    Have a nice day.

    <1 kiloflops

  50. meh... by beefguts · · Score: 1

    not much memory for that sized machine. Our M9000 machines have 2 TB of ram per node.

  51. Maintenance plan? by addikt10 · · Score: 1

    The eetimes article states a slightly more realistic 4096 processors per rack, or roughly 400,000 processors...

    Still, can you imagine the maintenance plan on this beast?
    Can you imagine the power and cooling involved?
    Even at only 25W per processor, we are talking nearly 10MW of power for the processors alone.

    Much more interesting than the machine itself would be an article on how they plan to keep it up and running.

  52. big enough for the next version of Windows? by peter303 · · Score: 0, Troll

    another obligatory post

  53. balanced computing: flops = memory words by peter303 · · Score: 1

    I recall this is some sort of named ad-hoc "law". When the amount core memory falls significantly below speeds, the kinds of computing you can do is severely limited. I believe they mainly plan simulations, where gigaflops per output point is typical and memory needs not as much. Data processing certainly desires balanced memory.

  54. Doesn't compute by mritunjai · · Score: 0, Redundant

    So 1.6M processors with 1.6TB RAM means just 1 MB RAM(1.6e+12/1.6e+6) per processor. That sounds bogus!

    Also roughly 12 GFLOPS of processing power per processor. WTF kind of cluster/super computer architecture is that ?? Sounds more like 1.6M Cell "stream" processors or something like that, definitely not something made from AMD/Intel parts. Of course, assuming numbers reported are correct.

    --
    - mritunjai
  55. Wait... by fataugie · · Score: 1

    Keep track of nuclear waste?
    A freakin pencil and paper wouldn't work for that?

    The rest of the duties are cool, more simulation and research and less underground testing...that's fine.
    But that initial reason is bogus!

    --

    WTF? Over?

  56. Cores NOT processors and... by Anonymous Coward · · Score: 0

    1.6 petabytes of memory

    quote
    "The Sequoia system will use 45nm processors with as many as 16 cores per chip running at a significantly faster data rate.

    Both BlueGene/P and Sequoia consist of clusters built up from 96 racks of systems. Sequoia will have 1.6 petabytes of memory feeding its 1.6 million cores, but many details of its design have not yet been disclosed. "

    I know I shouldn't be surprised at mistakes on slashdot, but the article is explicit about the details.

  57. Other purposes... by fahrbot-bot · · Score: 1

    Primarily to keep track of nuclear waste & simulate explosions of nuclear munitions, but also for research into astronomy, energy, the human genome, and climate change

    And determine the correct response to the question: "Does this dress make me look fat?"

    [Though it may need more processing power for that.]

    --
    It must have been something you assimilated. . . .
  58. Distributed computing gone obsolete. by Anonymous Coward · · Score: 0

    This puppy is probably going to find extraterrestrial life (seti) cure every disease known to man (f@h, rosetta etc) and still have an hour to spare before lunch.

  59. Total Information Awareness by Anonymous Coward · · Score: 0

    Hmm... Sounds perfect for processing heaps of telecommunications data.

  60. Too powerful to resist. by CobaltBlueDW · · Score: 1

    Is it just me or does petaflop sound like an impotent pervert?

    Well in any event, I'm sure with 1.6TB of memory it will be able to cache lots of petafiles.

  61. BS by More_Cowbell · · Score: 0, Redundant
    Sorry, need to call you out on that. You either suffer form some form of dyslexia or are karma whoring (for mods that don't RTFA). I actually was all set to mod you up, glad I really did RTFA.

    Boasting over 1.6 million processors and some 1.6TB of memory, the Sequoia system will initially be used to keep track of the US' mountain of aging nukes...

    Or were you saying that 'TB' is the new acronym for 'petabytes'?

    --
    Experience teaches only the teachable. -AH
    1. Re:BS by clodney · · Score: 1

      OK, now I am confused. The article that is currently in the summary is not the one I recall reading. And I specifically remember reading petabytes, not the abbreviation PB which I see in the informationweek article referenced in another comment.

      Color me confused, but not intentionally karma whoring, and not normally dyslexic.

    2. Re:BS by More_Cowbell · · Score: 1

      OK, my apologies. In my defense, the article linked in the summary said TB, as I noted...I've since read several others that say PB. (As a side note the others also said the number of CPUs listed actually refers to cores [which makes much more sense given the rack space noted]). Now I have no idea which article is correct, sorry for calling you out like that - it did seem clear cut at the time.
      If it makes you feel any better, your previous comment is still 'informative', while mine is 'redundant'. ;)

      --
      Experience teaches only the teachable. -AH
  62. Excel - The universal solution by Cassini2 · · Score: 1

    At one company I worked at, Excel was almost designated the company's standard word processor. You can write text in Excel. You can do tables and charts in Excel. You can do databases in Excel. Why use the other packages for anything else?

    The scary thing is, the proposal kind of makes sense ...

    1. Re:Excel - The universal solution by adolf · · Score: 1

      I'd guess that it's better than my company, where we're rather expected to install the same pirated[1] copy of Office 98 on every computer it is requested on. I, of course, refuse; others in the company aren't so clued.

      I've tried, over and over, to persuade them to either buy a bunch of copies of Office, or just stick with openoffice.org. It's not like they're using any advanced features of Excel or Word, or any parts (at all) of Access. Alas.

      [1]: I have mixed feelings about software piracy -- especially Microsoft software. Personallly: A lot of things I buy, a lot of things I don't. But if I'm using a thing in a business environment to make money, then, yes, the software should, absolutely, be paid for. I definitely bought every bit of software that I use in my sideline computer repair business -- I even register my shareware -- but unfortunately my employer doesn't want to operate that way.

  63. The real secret... by jonwil · · Score: 1

    Is the identical model IBM are building in a secret bunker underneath the NSA headquarters so the NSA can read everyones encrypted communications.

  64. Screen Saver by dazlari · · Score: 1

    but also for research into astronomy, energy, the human genome, and climate change.

    So they'll be running BOINC as a screen saver then?

  65. Re:What spy $ was ever well spent? by Anonymous Coward · · Score: 0

    Aren't you the guy that replies to his own posts? Why do you do that?

  66. Deep thought by Anonymous Coward · · Score: 0

    I can't help but notice completion of our "deep thought" coincides with the predicted (again) demise of our globe.

  67. Apocalypse by velja27 · · Score: 1

    On 2012 October IBM's supercomputer was launched. On 2012 November LHC was plugged into IBM's supercomputer. On 2012 December Black Hole destroyed Earth and its orbiting Moon.

  68. What will this be used for? by Anonymous Coward · · Score: 0

    You know what this will be used for, right? 1. Decrypt all email for surveillance. 2. Pattern match the internet for child porn. 3. Pattern match all the feeds from all cameras across the US for suspected terrorist activity. 4. Track down music piracy. etc. After all, these are all more important than tracking nuclear waste.

  69. Why the hell are we wasting all this power on nuke by melted · · Score: 1

    Why the hell are we wasting all this power on nukes? How about some advanced drugs? General AI? Complete simulation of human brain? I mean, Jesus H Christ, our nukes are already GOOD ENOUGH.

  70. Does it run COBOL? by pentrudownload · · Score: 1

    Does it run COBOL?

  71. "bailout" for politically connected IBM. by Anonymous Coward · · Score: 0

    As I have said in the past on this site... these huge computer clusters are jobs programs for politically connected IT corporations. These cluster projects keep assembly lines going for months. Look at pictures and words just a few days ago of a meeting between the CEO of IBM and Obama.

    The stated use for this computer system is absurd on it's face. One does not need a computer of this nature to perform the stated function.

    IBM does not deserve US government contracts. IBM has been enthusiastically engaged in economic sabotage by sending high paying technology jobs overseas. IBM is what it is because of the United States; and for the last decade IBM has been pissing on the United States and it's workers in the United States. Even if I hadn't witnessed IBM's treasonous behavior, I wouldn't want the US government to purchase an expensive computer it did not need.

  72. NNSA awards IBM contract by Anonymous Coward · · Score: 0

    With a speed of 20 petaflops Sequoia is expected to be the most powerful supercomputer in the world and will be approximately over 10 times faster than todayâ(TM)s most powerful system. To put this into perspective, if each of the 6.7 billion people on earth had a hand calculator and worked together on a calculation 24 hours per day, 365 days a year, it would take 320 years to do what Sequoia will do in one hour.

    https://publicaffairs.llnl.gov/news/news_releases/2009/NR-09-02-01p.html

  73. Yeah, tracking nuclear waste... by Anonymous Coward · · Score: 0

    It's called "cracking encrypted Internet messages concerning acquired nuclear waste for potentially nefarious activities". Seriously, it's a big encryption cracker. Give each CPU a series of keys to try and let her rip.

    I highly doubt anything related to science will grace this machine.

  74. How is this possible? by jon3k · · Score: 1

    1.6 million processors divided by 96 cabinets gives me 16,666 physical processors per cabinet. 16,666 processors / 42 rack mount units per rack gives me 396 processors per RMU.

    Either I'm missing something or they can fit 400 processors in 1U. Someone help me out here.

  75. Re:Why the hell are we wasting all this power on n by jon3k · · Score: 1

    To put it as simply as possible: so we don't have to actually detonate the nukes to test them.

  76. Re:Why the hell are we wasting all this power on n by melted · · Score: 1

    Why would you NEED to test them? They work already.

  77. Re:Why the hell are we wasting all this power on n by jon3k · · Score: 1

    Because they break down over time and we have to make sure they still work and will continue to work when we need them. They have components with half lifes that have to be monitored and replaced before they expire.