AMD EPYC Milan-X specifications leak, revealing up to 64 cores, 280W TDP, 768MB L3 cache.

Discussion in 'Frontpage news' started by Hilbert Hagedoorn, Sep 16, 2021.

  1. Hilbert Hagedoorn

    Hilbert Hagedoorn Don Vito Corleone Staff Member

    Messages:
    42,071
    Likes Received:
    10,076
    GPU:
    AMD | NVIDIA
  2. Denial

    Denial Ancient Guru

    Messages:
    13,527
    Likes Received:
    3,073
    GPU:
    EVGA RTX 3080
    I wonder if they are going to go the IBM route with virtual caches in the future. Seems like a much better solution for enterprise.
     
  3. icedman

    icedman Maha Guru

    Messages:
    1,124
    Likes Received:
    151
    GPU:
    MSI Duke GTX 1080
    I think amd is really onto something now with all the add on cache they're going to be throwing onto even their desktop CPU's
     
    tunejunky likes this.
  4. BLEH!

    BLEH! Ancient Guru

    Messages:
    6,116
    Likes Received:
    176
    GPU:
    Sapphire Fury
    That's a LOOOOOOT of cache! Is this like the L4 that intel stuck onto the desktop Broadwell parts then, or something else? I remember that giving a rather spicy boost to performance that kept the architecture on par with far more recent things...
     
    Noisiv and tunejunky like this.

  5. tunejunky

    tunejunky Ancient Guru

    Messages:
    1,644
    Likes Received:
    646
    GPU:
    RX 6800 RTX 2070
    yes for traditional enterprise you are probably right as long as "better" means cheaper.
    but for high tech enterprise nothing beats on die cache. and massive on die cache at that.
     
  6. nizzen

    nizzen Ancient Guru

    Messages:
    1,543
    Likes Received:
    467
    GPU:
    3x3090/3060ti/2080t
    Cache is king
     
    tunejunky likes this.
  7. Venix

    Venix Ancient Guru

    Messages:
    1,895
    Likes Received:
    807
    GPU:
    Palit 1060 6gb
    Hehe i guess we are not more than a year ot two to get cpus with 1gb cache ! Not on desktop thought :p
     
    tunejunky likes this.
  8. Denial

    Denial Ancient Guru

    Messages:
    13,527
    Likes Received:
    3,073
    GPU:
    EVGA RTX 3080
    It is on die - 256MB of Shared L2/L3 virtual cache per 8 core chip. It obviously comes with a latency penalty but the benefits of the L3 being shared with the L2 supposedly outweighs that.
     
  9. Kaarme

    Kaarme Ancient Guru

    Messages:
    2,619
    Likes Received:
    1,269
    GPU:
    Sapphire 390
    The latency is still, surely, an order of magnitude lower than accessing DDR?
     
    tunejunky likes this.
  10. user1

    user1 Ancient Guru

    Messages:
    1,733
    Likes Received:
    598
    GPU:
    hd 6870
    This is better, this is legit l3 cache, thats added on top of the existing cache and connected via tsvs (Through Silicon Via), there is no latency penalty, unlike an l4 cache which would be slower.
     
    Embra and tunejunky like this.

  11. tunejunky

    tunejunky Ancient Guru

    Messages:
    1,644
    Likes Received:
    646
    GPU:
    RX 6800 RTX 2070
    can you imagine (crazy sci - fi music) we're living in the fuuuutuuure
     
    Venix likes this.
  12. schmidtbag

    schmidtbag Ancient Guru

    Messages:
    6,147
    Likes Received:
    2,470
    GPU:
    HIS R9 290
    I'd be really curious to see what would happen if someone were to [somehow] put in a 64MB (not a typo) stick of RAM in, install Windows XP, and play some early 2000s game just to see how it runs. Obviously it's utterly useless, but it's weird to think about having more cache than you have RAM, and enough cache to actually run your entire system and a simple game. Hard to wrap your head around.
     
    tunejunky and BLEH! like this.
  13. BLEH!

    BLEH! Ancient Guru

    Messages:
    6,116
    Likes Received:
    176
    GPU:
    Sapphire Fury
    CacheRAMdrive? Cachedrive?
     
    schmidtbag likes this.
  14. Alessio1989

    Alessio1989 Ancient Guru

    Messages:
    2,112
    Likes Received:
    633
    GPU:
    .
    768 L3 cache

    [​IMG]
     
    tunejunky and Noisiv like this.
  15. Denial

    Denial Ancient Guru

    Messages:
    13,527
    Likes Received:
    3,073
    GPU:
    EVGA RTX 3080
    Yeah - the idea behind IBM's implementation is that it massively reduces L3 latency at the expense of slightly increased L2. If you get lucky, you can hit L3 for nearly identical latency as L2 but in the vast majority of scenarios the latency of your L3 is going to be significantly faster than traditional L3 - for example the L3 seen here. In traditional PC workloads you're not going to see much of a benefit but in use cases IBM's cloud servers, or Epyc's servers are being used, having a fast L3 is extremely useful.

    Keep in mind IBM's system is also scaling up to ~512mb of combined L2/L3 per module and 8GB of cache (virtual L4 in this case) in a 32 chip system.

    AMD's stacking here would be adjacent to this.
     
    Last edited: Sep 17, 2021
    tunejunky likes this.

  16. kapu

    kapu Ancient Guru

    Messages:
    4,852
    Likes Received:
    476
    GPU:
    Radeon 6800
    In before 1-2 gigs of cache and cache drives - can it run crysis from cache ?
     

Share This Page