• Scratch@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    5
    ·
    4 months ago

    <5% is within margin of error imo. Especially for a single run.

    Still! Even if losing by 5%, on a multi core test with 40% fewer cores is pretty nuts.

    • Valmond@lemmy.world
      link
      fedilink
      arrow-up
      1
      ·
      4 months ago

      5% spread in cpu power? Bet it’s lower than 0.5% or I’m missing something bizarre here.

      • Scratch@sh.itjust.works
        link
        fedilink
        English
        arrow-up
        1
        ·
        4 months ago

        I’m just trying to say that one point of reference shouldn’t be treated as hard fact. Especially if the results are close.

    • Alphane Moon@lemmy.worldOPM
      link
      fedilink
      arrow-up
      1
      arrow-down
      1
      ·
      4 months ago

      We definitely need some independent benchmarks on real world devices before making any kind of conclusions around “Strix Point”.

      That being said, some of the preview results do seem to suggest it’s not a complete dud and there may be a moderately decent offering here.