7 ms·

A bit more informative is the actual fact sheet put out by the white house [1]. What they are really aiming for is exascale computing, which they define as bein

by phreeza 11y ago

A bit more informative is the actual fact sheet put out by the white house [1]. What they are really aiming for is exascale computing, which they define as being capable of applying exaFlops to exabytes. From my limited knowledge, the latter will actually be the bigger deal. As pointed out elsewhere, an exaflop supercomputer will probably come around beforehand. [1] https://www.whitehouse.gov/sites/default/files/microsites/ostp/nsci_fact_sheet.pdf https://www.whitehouse.gov/sites/default/files/microsites/os...

ypcx 11y ago
Clearly designed to run AI for cyber-warfare / cyber-defense purposes? William Gibson's Black Ice coming to life?
- scottcha 11y ago
  More likely its for modeling nuclear weapons which I believe is the primary purpose of NSCI.
  - trsohmers 11y ago
    NNSA*, which is an office within the Department of Energy.
    - scottcha 11y ago
      National Strategic Computing Initiative (NSCI) is the initiative within the DOE.
  - semi-extrinsic 11y ago
    The story linked below about the French exascale plans explicitly states that it is the military applications division of the French nuclear energy agency (CEA) that is spearheading that effort. (I believe they work on both reactors and warheads.)
  - jen729w 11y ago
    Serious question: what's left to model? They're already big enough to kill everyone, and we have heaps of them.
    - markhahn 11y ago
      Often they model how the warhead ages, rather than how it explodes.
      - scottcha 11y ago
        Yes, this is my understanding.
trsohmers 11y ago
The real problem is getting 1 exaflop (or around it) within a reasonable power budget. The DOE's power budget for all of their supercomputing resources is 20 Megawatts, so at a full system level we would need to be at 50 GFLOPs per watt, while the best system right now is at 5.
- ghayes 11y ago
  > "I'd say they're targeting around 60 megawatts, I can't imagine they'll get below that," [Mark Parsons] commented. "That's at least £60m a year just on your electricity bill."
  - trsohmers 11y ago
    I wouldn't trust a Brit to report the figures for the US Department of Energy. The DOE's budget is 20 Megawatts [source: http://science.energy.gov/~/media/ascr/ascac/pdf/meetings/20140210/Top10reportFEB14.pdf http://science.energy.gov/~/media/ascr/ascac/pdf/meetings/20...]
    - andrewmu 11y ago
      He didn't report it, he gave an estimate. And as for "wouldn't trust a Brit", I'd trust anyone directly involved in it, regardless of nationality.
- tim333 11y ago
  Nvidia's Pascale GPU, due next year, is supposed to do 28 terraflops with 1300 watts which I make 21 GFLOPs per watt.
  - white-flame 11y ago
    20+ GFLOPS/W for a single part isn't something new: http://streamcomputing.eu/blog/2012-08-27/processors-that-can-do-20-gflops-watt/ http://streamcomputing.eu/blog/2012-08-27/processors-that-ca... (2012) A huge amount of overall system power is spent in data transport. Plus, double everything for cooling. That brings the total system efficiency way down from what the actual computational components spend.

by phreeza 11y ago

ypcx 11y ago

Clearly designed to run AI for cyber-warfare / cyber-defense purposes? William Gibson's Black Ice coming to life?

scottcha 11y ago
More likely its for modeling nuclear weapons which I believe is the primary purpose of NSCI.
- trsohmers 11y ago
  NNSA*, which is an office within the Department of Energy.
  - scottcha 11y ago
    National Strategic Computing Initiative (NSCI) is the initiative within the DOE.
- semi-extrinsic 11y ago
  The story linked below about the French exascale plans explicitly states that it is the military applications division of the French nuclear energy agency (CEA) that is spearheading that effort. (I believe they work on both reactors and warheads.)
- jen729w 11y ago
  Serious question: what's left to model? They're already big enough to kill everyone, and we have heaps of them.
  - markhahn 11y ago
    Often they model how the warhead ages, rather than how it explodes.
    - scottcha 11y ago
      Yes, this is my understanding.

trsohmers 11y ago

The real problem is getting 1 exaflop (or around it) within a reasonable power budget. The DOE's power budget for all of their supercomputing resources is 20 Megawatts, so at a full system level we would need to be at 50 GFLOPs per watt, while the best system right now is at 5.

ghayes 11y ago
> "I'd say they're targeting around 60 megawatts, I can't imagine they'll get below that," [Mark Parsons] commented. "That's at least £60m a year just on your electricity bill."
- trsohmers 11y ago
  I wouldn't trust a Brit to report the figures for the US Department of Energy. The DOE's budget is 20 Megawatts [source: http://science.energy.gov/~/media/ascr/ascac/pdf/meetings/20140210/Top10reportFEB14.pdf http://science.energy.gov/~/media/ascr/ascac/pdf/meetings/20...]
  - andrewmu 11y ago
    He didn't report it, he gave an estimate. And as for "wouldn't trust a Brit", I'd trust anyone directly involved in it, regardless of nationality.
tim333 11y ago
Nvidia's Pascale GPU, due next year, is supposed to do 28 terraflops with 1300 watts which I make 21 GFLOPs per watt.
- white-flame 11y ago
  20+ GFLOPS/W for a single part isn't something new: http://streamcomputing.eu/blog/2012-08-27/processors-that-can-do-20-gflops-watt/ http://streamcomputing.eu/blog/2012-08-27/processors-that-ca... (2012) A huge amount of overall system power is spent in data transport. Plus, double everything for cooling. That brings the total system efficiency way down from what the actual computational components spend.