Nvidia Ampere Discussion [2020-05-14]

Discussion in 'Architecture and Products' started by Man from Atlantis, May 14, 2020.

Tags:
  1. trinibwoy

    trinibwoy Meh
    Legend

    Joined:
    Mar 17, 2004
    Messages:
    11,813
    Likes Received:
    2,777
    Location:
    New York
    They're also marketing to the high refresh rate, low latency crowd though so you would think that efficiency at lower resolutions is also a priority.

    Pretty crap based on what I've seen so far but Nvidia said as much in the Ampere whitepaper. I've only looked at Cold War, Star Wars and Atomic Heart though. Overall SM utilization is usually somewhere around 20% during RT passes with the INT ALU pipe seeing lots of action.
     
    Cuthalu likes this.
  2. pjbliverpool

    pjbliverpool B3D Scallywag
    Legend

    Joined:
    May 8, 2005
    Messages:
    8,990
    Likes Received:
    3,792
    Location:
    Guess...
    That could bode well for the 3080Ti with less bandwidth but potentially very similar or even greater core performance than the 3090.
     
    PSman1700 likes this.
  3. trinibwoy

    trinibwoy Meh
    Legend

    Joined:
    Mar 17, 2004
    Messages:
    11,813
    Likes Received:
    2,777
    Location:
    New York
    I'm discounting the VRAM bandwidth stats as they don't really make sense. There's another metric "L2 bandwidth to VRAM" that is probably more representative. In some cases it's almost maxed out which would indicate a bandwidth bound situation.

    E.g. the HBAO pass in the Android mesh shader demo.

    [​IMG]
     
    T2098, jlippo, LeStoffer and 4 others like this.
  4. techuse

    Veteran Newcomer

    Joined:
    Feb 19, 2013
    Messages:
    1,062
    Likes Received:
    650
    Shades of FuryX. Bondrewd(I think that was his SN) was on point after all. With how power limited Ampere is I wonder if higher utilization levels would even net much of a performance win. They may just result in large clock reductions as an offset.
     
    ToTTenTranz likes this.
  5. Frenetic Pony

    Regular Newcomer

    Joined:
    Nov 12, 2011
    Messages:
    798
    Likes Received:
    463
    I don't even see how the rumor could point towards a consumer card. You'd need 4 stacks of HBM 2e for relative bandwidth there, that's just plain expensive. And as pointed out, it's not like anyone's 5nm is some magical node that halves power usage either. Are they going to try and make a 700 watt card? Quad slotted propelyne glycol/water mix cooler?
     
    HLJ likes this.
  6. Man from Atlantis

    Regular

    Joined:
    Jul 31, 2010
    Messages:
    942
    Likes Received:
    804
    I can't find my old nsight installer and had to install the latest version on a fresh windows. It's much more detailed now but it's surprise to see there's no FP16+Tensor pipe throughput anymore and FMA pipe throughput's divived to FMALite and FMAHeavy.

    Can any Turing owner confirm if it's still show as FP16+Tensor pipeline on their nsight?

    [​IMG]

    https://docs.nvidia.com/nsight-compute/ProfilingGuide/index.html
     
  7. sniffy

    Newcomer

    Joined:
    Nov 14, 2014
    Messages:
    55
    Likes Received:
    83
    pharma likes this.
  8. Man from Atlantis

    Regular

    Joined:
    Jul 31, 2010
    Messages:
    942
    Likes Received:
    804
    It's Nsight Graphics, but the excerpt is only available on the Nsight Compute documentation. Probably haven't updated on Nsight graphics documantation yet.
     
    #2588 Man from Atlantis, Dec 29, 2020
    Last edited: Dec 29, 2020
  9. trinibwoy

    trinibwoy Meh
    Legend

    Joined:
    Mar 17, 2004
    Messages:
    11,813
    Likes Received:
    2,777
    Location:
    New York
    Yeah the documentation of the individual metrics in Nsight graphics is really poor. I also just noticed that there’s a “pro” build of Nsight that shows RT core activity that’s not available in the public version. Sucks.
     
    Scott_Arm, BRiT and Man from Atlantis like this.
  10. Man from Atlantis

    Regular

    Joined:
    Jul 31, 2010
    Messages:
    942
    Likes Received:
    804
    It's sad, it used to show DXR dispatch and DXR build sections under DXR marker at the bottom.

    In the excerpt it says;
    Does that mean FMA heavy and FMA lite have physically different pipelines, like INT, FP64, Tensor ops on Ampere?
     
  11. trinibwoy

    trinibwoy Meh
    Legend

    Joined:
    Mar 17, 2004
    Messages:
    11,813
    Likes Received:
    2,777
    Location:
    New York
    It still shows the BVH build and ray dispatch in the markers section but there’s no separate DXR row any more and no way to see RT core utilization. I’m not sure if the old DXR row was just a marker or actual throughput stats.

    FMA heavy seems to count FP instructions running on the FP+INT pipe on Ampere while FMA lite is the FP only pipe. I don’t see anything for Tensors in Nsight any more even when running DLSS.
     
    Scott_Arm and Man from Atlantis like this.
  12. pharma

    Veteran Regular

    Joined:
    Mar 29, 2004
    Messages:
    4,642
    Likes Received:
    4,125
    They do have prior nsight versions available to download if you recall the version number.

    There's a pro version? I thought the only version available was the one you got once registered.[/QUOTE]
     
    Man from Atlantis likes this.
  13. Man from Atlantis

    Regular

    Joined:
    Jul 31, 2010
    Messages:
    942
    Likes Received:
    804
    Thanks, unfortunately versions prior to 2020.5 don't support Ampere GPUs.
     
    pharma likes this.
  14. trinibwoy

    trinibwoy Meh
    Legend

    Joined:
    Mar 17, 2004
    Messages:
    11,813
    Likes Received:
    2,777
    Location:
    New York
    "For users who have access to the Pro builds of Nsight Graphics"

    Notice some metrics have (Pro) next to them.

    User Guide :: Nsight Graphics Documentation (nvidia.com)
     
    BRiT and pharma like this.
  15. pharma

    Veteran Regular

    Joined:
    Mar 29, 2004
    Messages:
    4,642
    Likes Received:
    4,125
    If you substitute the version number in the link other archived versions are available.
    i.e.
    https://developer.nvidia.com/nsight-graphics-2020_6
     
  16. Man from Atlantis

    Regular

    Joined:
    Jul 31, 2010
    Messages:
    942
    Likes Received:
    804
    I've tried both 2020-05 and 2020-06 already, these are the ones that support Ampere lineup. Prior version don't support. I wonder if the latest version gives different SM Instruction Throughput readup for Turing lineup, like still showing FP16+Tensor and/or unified FMA pipe throughput.
     
    pharma likes this.
  17. DmitryKo

    Regular

    Joined:
    Feb 26, 2002
    Messages:
    967
    Likes Received:
    1,222
    Location:
    55°38′33″ N, 37°28′37″ E
    Rumored to have 18432 CUDA cores and 64 TFLOPS... 2022 is going to be an interesting year.


    BTW the articles also list other NVidia codenames (supposedly leaked in the 2018 GPU Technology Conference keynote slide) which serve as tributes to renowed physicists and computer scientists:
    Turing has already been used, while Lovelace and Hopper are rumored for 2022; the remaining 6 codenames should probably suffice until 2030 (so far they've only used 8 codenames since 2006 - Tesla, Fermi, Kepler, Maxwell, Pascal, Volta, Turing, Ampere).

    [​IMG]
     
    #2597 DmitryKo, Dec 30, 2020
    Last edited: Dec 31, 2020
    PSman1700, pharma and sonen like this.
  18. PSman1700

    Legend Newcomer

    Joined:
    Mar 22, 2019
    Messages:
    5,839
    Likes Received:
    2,534
  19. cho

    cho
    Regular

    Joined:
    Feb 9, 2002
    Messages:
    422
    Likes Received:
    16
    NVIDIA told me the pro version nsight graphics that can show RT core usage require a company NDA.
     
    Lightman, Kej, Scott_Arm and 4 others like this.
  20. trinibwoy

    trinibwoy Meh
    Legend

    Joined:
    Mar 17, 2004
    Messages:
    11,813
    Likes Received:
    2,777
    Location:
    New York
    Interesting. Wonder what’s so sensitive about those stats.
     
    LeStoffer, Krteq and nutball like this.
Loading...

Share This Page

  • About Us

    Beyond3D has been around for over a decade and prides itself on being the best place on the web for in-depth, technically-driven discussion and analysis of 3D graphics hardware. If you love pixels and transistors, you've come to the right place!

    Beyond3D is proudly published by GPU Tools Ltd.
Loading...