Comments Locked

27 Comments

Back to Article

  • Drumsticks - Tuesday, June 20, 2017 - link

    The V100 is an incredible part. I wonder just how many will actually be available though. It can't be easy to manufacture something over 800mm area.

    I also wonder how this will lead into Volta having cards. The P100 was faster than the initial run off Pascal. I wonder if we will actually be seeing the performance flag being moved next year, or if a small card like the 750 ti, or just a more efficient flagship. Think 1080 Ti, but less power. Maybe Vega won't actually have to compete with a GV102 until late 2018ish.
  • Qwertilot - Tuesday, June 20, 2017 - link

    Best guess - and it obviously is a real guess, but they are very predictable - is a repeat of what they did with Pascal. So an 1170/80 with the 1180 a little ahead of the 1080ti in performance while using rather less power.

    Then a Volta titan (V102 without all the compute stuff) with an 1180ti coming later.
  • Drumsticks - Tuesday, June 20, 2017 - link

    That's a fair point. The last time we had an arch revisdion without a major node change (780 Ti to 980), absolute performance didn't move a ton, maybe 10 to 20%, but obviously efficiency was way up. I wonder what we'll be seeing from Volta on the gaming side. From a TF efficiency perspective, the 980 was about 400GF below the 780 Ti, but about 10-30% faster. It'll be interesting to see if they can move that mark again with Volta.
  • FreckledTrout - Tuesday, June 20, 2017 - link

    With what they will charge 800mm wont matter even if they end having to toss half them in the trash. The DGX Station with 4 of these was quoted to cost $69,000.
  • andychow - Wednesday, June 28, 2017 - link

    I was just about to ask: How are they making a 815mm2 chip? I thought the max was 700mm2, on Intel special fabs, and TSMC GloBo could do 610mm2 max. I'm not even talking about defects, but physical limitations like reticle size. Where are these being manufactured?
  • stanleyipkiss - Tuesday, June 20, 2017 - link

    How are the passively cooling a 250 W card?
  • nathanddrews - Tuesday, June 20, 2017 - link

    Because the rack environment such a card is used in sounds like an aircraft runway.
  • ddriver - Tuesday, June 20, 2017 - link

    There are fans on the rackmount case that pull air through the entire chassis.
  • kaesden - Tuesday, June 20, 2017 - link

    its not really passively cooled when installed in a system. The card itself has no active cooling elements, but its put into an environment with loads of airflow from the server chassis. Servers don't give a f- about noise.
  • Gothmoth - Tuesday, June 20, 2017 - link

    nvidia should start producing x86 cpus too.

    i know i know.. it is not that easy blah blah.. just saying that i would find it nice to have more competition.
  • bubblyboo - Tuesday, June 20, 2017 - link

    Back in 2011 Intel and Nvidia settled over patents and the consequences for Nvidia was that they are pretty much NEVER going to get an x86 license or be able to make x86 emulators.
    http://www.anandtech.com/show/4122/intel-settles-w...
  • Yojimbo - Tuesday, June 20, 2017 - link

    The settlement didn't grant NVIDIA any x86 license, as in it didn't force Intel to provide NVIDIA with one, but it doesn't bar Intel from granting NVIDIA an x86 license. NVIDIA won't get an x86 license because Intel doesn't want them to, not because of the settlement agreement.
  • willis936 - Wednesday, June 21, 2017 - link

    And then have a billion dollar antitrust lawsuit brought against them.
  • andychow - Monday, September 18, 2017 - link

    The earliest patents expiration for basic x86-64 implementation is 2025. They need to have legal access to things like US6877084, and that will never happen until the patent expires. They could do basic 586 type processors, but they would only be 32-bit, 2GB ram, and not run on most modern OS, which don't even support 32-bit anymore.
  • vFunct - Tuesday, June 20, 2017 - link

    Any guess on price? $10,000? $25,000?
  • Morawka - Tuesday, June 20, 2017 - link

    i'd venture to say 15k each
  • Bateluer - Tuesday, June 20, 2017 - link

    An earlier AT article on the GV100 Volta cards stated 18K/card.
  • Ryan Smith - Tuesday, June 20, 2017 - link

    Note that 18K was the effective price you were paying since you had to buy a DGX-1 to get them. It is not the stand-alone price.
  • bill.rookard - Tuesday, June 20, 2017 - link

    Folding rig anyone? Anyone? Sounds like it's got a stupid amount of processing power which is pretty darn awesome.
  • Flunk - Tuesday, June 20, 2017 - link

    Sure, have a spare $20,000/card?
  • edzieba - Wednesday, June 21, 2017 - link

    "Interestingly, unlike the Tesla P100 family, NVIDIA isn't offering a second-tier PCIe card based on salvaged chips; so this generation doesn't have an equivalent to the 12GB PCIe Tesla P100. NVIDIA's experience with GP100/interposer/HBM2 assembly as well as continuing production of HBM2 has likely reduced the need for memory-salvaged parts."

    That is VERY interesting. While rampant speculation, the other possibility is that salvaged assemblies are instead being stockpiled for a Titan-eqsue (or actual Titan) more consumer-oriented card. With the difficulty everyone is experiencing assembling HBM 1 & 2 packages, having such a low defect rate that failed packages can just be discarded seems less likely then collecting them for a lower-margin 'halo' consumer card.
  • jabbadap - Wednesday, June 21, 2017 - link

    If I must guess, probably quadro line. Why offer something at $1200 when you can ask $6000 for it with the right name on it.
  • frenchy_2001 - Friday, June 23, 2017 - link

    Even for workstation the Big Chips make little sense.
    P100 and now V100 are heavily compute oriented. If it follows Pascal (the Quadro P6000 was based on GP102, same as the Titan card and 1080Ti), workstation will go with V102.
    V100 is a monster for AI (particularly tensor operations). We'll have to wait a few months to see if they introduce a memory cut compute card or a lower class (workstation/gaming).
  • MadManMark - Wednesday, June 28, 2017 - link

    That makes literally no sense. You are suggesting they're choosing to forgo a higher margin enterprise card in order to make less money (or even lose money?) on a consumer card.
  • Ahnilated - Wednesday, June 21, 2017 - link

    Is anyone else getting tired of these pre release releases? They don't have product and aren't available for months but have to get some information out. I am so tired of this stuff it isn't funny anymore.
  • K_Space - Thursday, June 22, 2017 - link

    Probably because of AMD Vega Frontier Edition, Nvidia is alerting potential buyers to hold fire?
    Any noticed how AMD plays on Nvidia's wording? Vega/Volta ... Founders Edition/Frontier Edition?
  • beck2050 - Saturday, June 24, 2017 - link

    These should provide a massive leap in performance. Excited to see gaming version. It seems like Nvidia is accelerating release times as not that long ago everyone was saying no Volta until 2018.

Log in

Don't have an account? Sign up now