Comments Locked

27 Comments

Back to Article

  • Mobile-Dom - Tuesday, August 11, 2015 - link

    see i like the M4000, is there a particular reason that the 970 couldn't have been a single slot card if the M4000 is the closest analogue?
  • close - Tuesday, August 11, 2015 - link

    Waiting for Nvidia to pull out of their hat a workstation card that has FP64 1/256. We like big numbers even when they're on the denominator...
  • Yojimbo - Tuesday, August 11, 2015 - link

    Yes, why not?
  • MrSpadge - Tuesday, August 11, 2015 - link

    1/32 or 1/256 almost makes no difference. If you need FP64 you need one of the big cards anyway. Otherwise there are plenty of use cases for FP32 focussed cards.
  • ddriver - Tuesday, August 11, 2015 - link

    Way to go, 8 times the difference is equal to almost no difference. I guess you almost won't care if your boss cuts your salary 8 times right?
  • univern72 - Tuesday, August 11, 2015 - link

    Right! If my salary were $8 trillion and my boss cut it to an 1/8th of that, I almost wouldn't care.
  • close - Wednesday, August 12, 2015 - link

    Maxwell sucks for HPC. Big Maxwell is 1/32 while big Fermi/Kepler were 1/2 and 1/3 and big Tesla is 1/8.

    And when you say "makes no difference" do you mean it or just meant "it's already so bad that it's like being in sh*t up to your waist or up to your bellybutton"?
  • testbug00 - Tuesday, August 11, 2015 - link

    er, correct me if I'm wrong, but, 1/32 is the lowest it gets when doing it in software unless you mess it up on purpose?
  • ddriver - Tuesday, August 11, 2015 - link

    Better get back to the roots, and only have full throughput on 16 bit precision. Like the good old voodoo. Professionals don't need neither good FP64 performance, nor even good FP32 performance for that matter, so go ahead and launch next gen GPUs with 1/64 the FP32 and 1/1024 FP64 performance.
  • LukaP - Wednesday, August 12, 2015 - link

    actually for the kind of work these cards are aimed at (namely rendering videos and 3D) you dont need FP64. basically every renderer is using FP32 anyway. where FP64 comes into play is with the teslas and there i am sure nvidia will come up with a derivation of maxwell with added 64 bit SIMD units
  • edzieba - Tuesday, August 11, 2015 - link

    Probably because the M4000 is boosting to 780MHz, whereas the 970 is hitting 1250MHz. Added to that the GM204 in the M4000 is going to be a more stringently binned part than those found in the 970, and thus more tolerant of continued high temperatures. If you slapped the M4000 cooler on a 970, it would be thermally throttling in short order.

    Would I like a 970 'lite' with that slim cooler? Heck yes! But the market for single-slot GPUs isn't huge at the moment, and OEMs are focussed on marketing 'short' GPUs under the ITX banner (even though those end up being taller and thicket than useful in most cases). Cases using rotated GPUs, either through edge-projecting PCIe slots (like on the B85M-VIEW PAKER) or flex risers for a 'folded path', can be made significantly thinner with single-slot cards, but are a niche market.
  • testbug00 - Tuesday, August 11, 2015 - link

    Don't forget the memory! Clocking it at "only" 6Ghz instead of 7Ghz saves quite a bit of power. :)
  • dragonsqrrl - Wednesday, August 12, 2015 - link

    T.D.P.

    In this sense the 970 isn't the closest analog, the 960 is. And even at 120W your probably pushing the thermal limit for a single slot card, which is probably why you don't see any single slot 960's either. There is a single slot 750 Ti, but that's 60W.
  • Jaguar36 - Tuesday, August 11, 2015 - link

    Real or Rendered is such a silly concept for a video card maker. That all has to do with software, not with the hardware. Its going to look just as good on their cheapest card, or even on an intel integrated graphics. Just going to take a heck of alot longer to render.
  • kaidenshi - Tuesday, August 11, 2015 - link

    I think the insinuation is just that: You can get photorealistic renders in a fraction of the time compared to integrated or non-Nvidia graphics cards. I think they are also trying to show off features they believe are unique to their software and drivers.
  • ddriver - Tuesday, August 11, 2015 - link

    Yeah, or get a 3d party OpenCL renderer and a radeon which will compeltely and utterly annihilate the quadro, not only in terms of raw FP64 performance, but even more so in price/performance ratio.
  • jwcalla - Tuesday, August 11, 2015 - link

    Ummm... is FP64 even used in rendering?

    Does AMD have their OpenCL stuff working in Blender yet, or is that still a trainwreck?
  • LukaP - Wednesday, August 12, 2015 - link

    It is not (or very rarely, definitely never when youre thinking about which card is faster and comparing to a radeon)
  • Yojimbo - Tuesday, August 11, 2015 - link

    NVIDIA does not make video cards, and I believe they have more software engineers than hardware engineers.
  • testbug00 - Tuesday, August 11, 2015 - link

    Nvidia DOES make video cards. They don't make many. Mostly they make silicon and require AIBs to build things to a minimum spec.

    Their victories and profitability pretty much all comes from their software however, as their GPUs (speaking on hardware) have typically been worse and won due to having good software versus AMD having not-as-good software. Recently they have taken a lead in power consumption and not compute tasks with their physical designs however.
  • Gadgety - Tuesday, August 11, 2015 - link

    So how does Desginworks position Nvidia iin relation to other render engines?
  • boeush - Tuesday, August 11, 2015 - link

    I'm not impressed with the drill. Show me a realistic teacup, or go home... ;p
  • boeush - Tuesday, August 11, 2015 - link

    I meant teapot... But a teacup to go with it would be nice too.
  • yannigr2 - Tuesday, August 11, 2015 - link

    Real or rendered?

    You could ask that question from the times of Amiga computers.
  • yannigr2 - Tuesday, August 11, 2015 - link

    M4000 specs I guess go like this

    7+1 GDDR5
    224+32bit data bus
    56 ROPs
  • huaxshin - Tuesday, August 11, 2015 - link

    Leaks of Quadro M5000 found here. Cheers ;)
    (Mobile M5000)

    http://forum.notebookreview.com/threads/17-4k-3840...
  • adithyay328 - Wednesday, August 26, 2015 - link

    Looks cool. They've got a new SFF quadro, which could prove useful for mini-itx users who want that power.

Log in

Don't have an account? Sign up now