Announcement

Collapse
No announcement yet.

AMD Sends Out Linux Graphics Driver Patches For "Arcturus" As New Vega Derived GPU

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #11
    Originally posted by ihatemichael
    With all due respect: I don't like having to set up different environment variables for my desktop to work correctly. While I appreciate the fact that this helped, it still feels a lot like a workaround and not a real fix.
    OK, understood. Just checking though - do the artifacts still go away if you do set the environment variable ?

    Depending on whether other users are seeing the same problem it may make sense to change the default in upstream code... I don't *think* I have seen this as a common problem on Raven but Marek might know better.

    Test signature

    Comment


    • #12
      Putting on my speculation hat, and having glanced over available reports and through the patches, it looks very much like a widened Vega 20 -- the reported 8192 SPs and 2x cache size, the 'MI-100' callsign, wide-loads, and what seem to be indications of 2x wider SIMD registers all point that way. So how are they doing it? I don't think its a dual-package or dual-die with InfinityFabric in between, or I don't think we'd see all this 2x widening going on; and besides that Vega 20 is only ~330mm square with graphics function intact, doubling the SPs while dropping graphics functions would probably come in comfortably under 600mm square, and much smaller than nVidia's reticle-busting ~800mm square dies.

      I also believe more recent versions didn't so much hit the 4096 SP limit due hard architecture limits, but moreso with the difficulty in balancing them against the practical limits of adding more graphics-related things around them, in balance. That idea is bolstered by the fact that available GCN projects have scaled far better with compute workloads than with rendering, and perhaps also that memory overclocking has been more profitable than core overclocking for rendering workloads.

      The simplest idea might be to just double the SIMD, register-width, and datapath across the board, which would seem to be straight-forward once freed of concerning itself with rendering workloads.

      But I also wonder if compute HPC workloads would benefit at all from more ACEs and if their might be some vestigial architecture limits related to that... So looking at how Navi is organized, I wonder if they're doing the dual shader-engine thing with 64 x 2 shader engines -- I believe the fine-grained scheduling at the shader engine level already has enough work available to keep two independent groups of SIMDs busy.

      Now, what would be really cool is if this were a not just a socketed product, but if there were also CPU cores in the package or on-die along with a full compliment of DDR4 channels in addition to on-package HBM acting as cache or scratchpad. It seems to me that some number of sockets all filled with those would make a more-regular HPC architecture than one where you have the same number of sockets where one-out-of-N is the odd man out. If you look at some of the other vector architectures out there, they have fairly strong scalar CPUs alongside the vector engines -- not ~8-wide super-scalars like Ryzen or Skylake, but more than the scalar capabilities inside GPU architectures.

      Comment


      • #13
        Originally posted by bridgman View Post
        ...
        Is the Radeon VII officially EOL? If so, can you tell Lisa that we need a replacement for compute for the people who cannot fit full supercomputers in their basement or attic? Thanks ;-)

        Comment


        • #14
          Originally posted by hoohoo View Post
          Is the Radeon VII officially EOL? If so, can you tell Lisa that we need a replacement for compute for the people who cannot fit full supercomputers in their basement or attic? Thanks ;-)
          No idea, sorry... that's about 85 departments away from us.

          Guessing you mean "not building more" rather than actual EOL ? I see the term "EOL" being used a lot but AFAIK that means ending support, not ending production.

          If/when I hear about production stopping I will make sure the question of "what are we recommending for entry level FP64 compute now ?" gets asked.
          Last edited by bridgman; 19 July 2019, 01:06 PM.
          Test signature

          Comment


          • #15
            Originally posted by bridgman View Post

            No idea, sorry... that's about 85 departments away from us.

            Guessing you mean "not building more" rather than actual EOL ? I see the term "EOL" being used a lot but AFAIK that means ending support, not ending production.

            If/when I hear about production stopping I will make sure the question of "what are we recommending for entry level FP64 compute now ?" gets asked.
            Thanks!

            (I did mean "not building more", I was speaking imprecisely.)

            Comment


            • #16
              Originally posted by hoohoo View Post
              Is the Radeon VII officially EOL?
              It's probably superseded by the Radeon Pro Vega II, kinda like how their Vega Frontier Edition was superseded by the Radeon Pro WX 9100.

              Comment


              • #17
                Originally posted by coder View Post
                It's probably superseded by the Radeon Pro Vega II, kinda like how their Vega Frontier Edition was superseded by the Radeon Pro WX 9100.
                The Radeon Pro Vega II is a more powerful videocard, certainly. But it uses a non-standard host interface and is only available for one Apple computer model. I'm just a PC, if you will. If it doesn't use standard PCIe3 or 4 then I can't consider it.

                Comment


                • #18
                  Originally posted by hoohoo View Post

                  The Radeon Pro Vega II is a more powerful videocard, certainly. But it uses a non-standard host interface and is only available for one Apple computer model. I'm just a PC, if you will. If it doesn't use standard PCIe3 or 4 then I can't consider it.

                  Comment


                  • #19
                    Originally posted by hoohoo View Post
                    The Radeon Pro Vega II is a more powerful videocard, certainly. But it uses a non-standard host interface and is only available for one Apple computer model.
                    It looks to me like it has a normal PCIe connector and a second connector for Thunderbolt. Then, there's the over-the-top connector, for Infinity Fabric. I'll bet it plugs into a PC just fine. Of course, that doesn't help if you can't buy them on the open market.

                    Originally posted by hoohoo View Post
                    I'm just a PC, if you will. If it doesn't use standard PCIe3 or 4 then I can't consider it.
                    I expect they'll release a PC-only version without the Thunderbolt connector, in the near future. They're probably just holding back to ensure adequate supply for Apple.

                    Comment


                    • #20
                      Originally posted by coder View Post
                      It looks to me like it has a normal PCIe connector and a second connector for Thunderbolt. Then, there's the over-the-top connector, for Infinity Fabric. I'll bet it plugs into a PC just fine. Of course, that doesn't help if you can't buy them on the open market.


                      I expect they'll release a PC-only version without the Thunderbolt connector, in the near future. They're probably just holding back to ensure adequate supply for Apple.
                      The second connector physically would intersect with various bit & pieces of an ATX mobo, though I expect you are correct at the electrical level.

                      I'd be interested in acquiring a PC compatible dual GPU Radeon VII card with 32+ GB HBM2 if AMD makes one (hint, hint @bridgman).

                      Comment

                      Working...
                      X