Fire in the hole!

  • mojofrododojo@lemmy.world
    link
    fedilink
    English
    arrow-up
    4
    arrow-down
    1
    ·
    2 days ago

    oh yeah running a rack of gpus full out for inference and whatnot is exactly like playing some games.

    buuuullll SHIT.

    • potoooooooo 🥔@lemmy.world
      link
      fedilink
      English
      arrow-up
      4
      arrow-down
      1
      ·
      2 days ago

      Also, tons of people have complained about the increasing power requirements of increasingly large GPUs, the increasing hardware demands of “AAA” games despite the low quality, etc… An entire universe of indie games has emerged in part because of that bullshit.

      • mojofrododojo@lemmy.world
        link
        fedilink
        English
        arrow-up
        2
        arrow-down
        1
        ·
        2 days ago

        the increasing power requirements of increasingly large GPUs

        yup - a friend of mine has a dead 3080 - nvidia did not do a good job designing those power connectors.

      • The_Terrible_Humbaba@slrpnk.net
        link
        fedilink
        arrow-up
        1
        arrow-down
        2
        ·
        2 days ago

        “Tons of people”

        I was one of those people, on Lemmy, and I’m telling you first hand the vast majority of people did not give a shit, they just wanted a more powerful GPU

      • mojofrododojo@lemmy.world
        link
        fedilink
        English
        arrow-up
        3
        ·
        2 days ago

        local morals

        lol. I know you mean local models, but also I suspect your local morals are lacking too considering the other comment.

        the inference and model construction take grunt bud; just because you’re not doing it yourself doesn’t make it free.

          • mojofrododojo@lemmy.world
            link
            fedilink
            English
            arrow-up
            2
            arrow-down
            1
            ·
            1 day ago

            No it doesn’t,

            the inference and model construction certainly do take more grunt, why are you fucking lying? what model do you use that was rolled on a home gpu?

            this is why I can’t stand you AI shills, you’ll create a bunch of strawman distractions, ignore the actual threats the tech poses, then pout when people don’t believe your lies.