AMD CDNA Discussion Thread

Discussion in 'Architecture and Products' started by Frenetic Pony, Nov 16, 2020.

  1. Leoneazzurro5

    Leoneazzurro5 Regular

    Trento + MI200?
     
  2. Bondrewd

    Bondrewd Veteran

    Yea.
     
  3. Bondrewd

    Bondrewd Veteran

  4. Krteq

    Krteq Newcomer

  5. DegustatoR

    DegustatoR Veteran



    [​IMG]
     
  6. Bondrewd

    Bondrewd Veteran

    Yes, that lidded 3x reticle CoWoS option for HPC was made for MI200s.
     
    Lightman and Krteq like this.
  7. Bondrewd

    Bondrewd Veteran

    boom
     
    Lightman and CarstenS like this.
  8. Leoneazzurro5

    Leoneazzurro5 Regular

    N6 process.
     
    Lightman likes this.
  9. Bondrewd

    Bondrewd Veteran

    All AMD parts are N6 now
     
    Krteq, BRiT and Lightman like this.
  10. Bondrewd

    Bondrewd Veteran

    AWWWWWWWWWWW YIS IT'S THE FIRST -LSI part

    [​IMG]

    Or maybe it's an ASE part.
    Time to annoy people.
     
    Last edited: Nov 8, 2021
  11. Granath

    Granath Newcomer

    no cowos? EFB.
     
  12. Bondrewd

    Bondrewd Veteran

  13. DavidGraham

    DavidGraham Veteran

    This isn't a one single wholesome GPU though, It's a two die GPU and appears as two GPUs to the system, juggling data between the dies is way slower than a single monolithic die, as the link between the two dies can only handle 200GB/s between each die, so aggregate performance numbers are theoretical numbers and should not be taken at face value.

    Also, on contrast to NVIDIA's A100, this isn't primarily a matrix/AI GPU, but rather is a vector GPU primarily. As a result, AI enhancements are very modest compared to A100 even when combining the two dies together .. however, vector throughput is through the roof. AMD and NVIDIA is diverging here, AMD is still not fighting NVIDIA in the AI market, rather focusing solely on the HPC market.

    Power consumption is 560W and is liquid cooled, which is curious considering the 6nm process.
     
    xpea, PSman1700 and pharma like this.
  14. Bondrewd

    Bondrewd Veteran

    That's all DC GPUs going forward due to NUMA.
    A100 programming in some cases is already that due to wonky as balls split L2.
    More, like 400.
    No, this thing is just purpose-built.
    They'll make a blingee matrix engine when the wider market wants their stuff.
    500, 560 is the actual Frontier SKU as per stuffed into that Shasta blade.
     
    Lightman likes this.
  15. OlegSH

    OlegSH Regular

    https://www.amd.com/en/graphics/server-accelerators-benchmarks
    AMD benchmarks.

    On average MI250 is up to 2.3x faster in comparison with A100 in DP workloads with 1 GPU config mostly (except for LAMMPS).
    The largest gains are in LINPACK of course.
    A100 still has an edge in GPU to GPU communication bandwidth, so the difference with real multi-GPU configs will likely be smaller.
     
    PSman1700, xpea, DavidGraham and 2 others like this.
  16. pharma

    pharma Veteran

  17. MI200 series is old.
    Cool folks are talking about MI300 already

    Probably for delivery to El Cap
     
    Man from Atlantis and Lightman like this.
  18. Bondrewd

    Bondrewd Veteran

    Yessss.
    Obviously, big DOE heads quoted the El Cap timeframe already.
    It's essentially 5Q after Frontier.
     
    Lightman likes this.
  19. I wonder if they will revise the perf target given latest exascale showdown with far east contenders.
    Already saw some govt. head honchos on twitter debating.
     
  20. Bondrewd

    Bondrewd Veteran

    El Cap is like puny 33MW so throwing moar nodes at the problem is totally feasible.
    It's a closed off literally simulate nukes system so do they really have to?
     
Loading...

Share This Page

Loading...