• ben@lemmy.zip
    link
    fedilink
    English
    arrow-up
    60
    ·
    11 months ago

    Took long enough - at a certain point Nvidia’s pricing just to get CUDA doesn’t make sense when compared to the cost of just investing in ROCm and OneAPI.

    All they had to do was find the right balance, but apparently they decided to see how much money the printer could make…

    • realharo@lemm.ee
      link
      fedilink
      English
      arrow-up
      22
      ·
      11 months ago

      I keep hearing how good AI is at coding these days, why can’t they just use it to rewrite all the model and library code up to full AMD support?

      /s