• @kadu
    link
    English
    91 year ago

    They desperately need to include hardware dedicated to real time AI upscaling.

    • @sheogorath
      link
      English
      231 year ago

      AMD simply doesn’t have the R&D capability to overtake NVIDIA on the technology front right now. When developing Ryzen, they took advantage of Intel getting complacent on all fronts. AMD also had Jim Keller in laying the groundwork for Zen architecture.

      However with NVIDIA, even though the consumer side offering has been underwhelming, they still kept going full throttle on the development of their enterprise parts. So if AMD ever got close enough on the performance side, NVIDIA only need to trickle down their chip design from enterprise/HPC to the consumer side.

      So, if AMD really want to take market share from NVIDIA on the consumer front, they’ll need to undercut on pricing real hard. Like offering 80%-90% of the performance for 60% of the price. But from what I’ve seen from AMD behavior for the 7900XTX release, they’re content with the status quo as it is.

      The only wildcard in GPU space right now is Intel. I sincerely hope they can light a fire on AMD and NVIDIA asses and start a performance wars that we’ve enjoyed in CPU space since Zen 2 got released.

      • @[email protected]
        link
        fedilink
        English
        31 year ago

        Nvidia just has been the dominant player in workstation GPUs for so long that it’s almost impossible to see AMD catching up. The hardware is one thing, but I don’t see AMD closing the gap even if they had a GPU that could perform as well as Nvidia’s. Just look at some of the benchmarks on stuff like blender and unreal engine. And that’s not to mention all of the work that’s gone on for optimizing ML and deep learning models for CUDA.

        I won’t pretend like I’m any sort of expert on GPUs, but I feel that Intel and AMD need to team up to create some standard software implementation for AI/ML, and then support it instead of just relying on open source (like with Pytorch and ROCm)

    • @UnaSolaEstrellaLibre
      link
      English
      51 year ago

      They already have some matrix accelerators in RDNA3 GPUs. As for if they’ll be used for something like FSR3 nobody knows.