• Hypx@fedia.io
    link
    fedilink
    arrow-up
    5
    arrow-down
    2
    ·
    3 months ago

    It’s the 12VHPWR connector that’s melting. The problem is that it is much smaller than the connector it replaces, while also sending much more power. Without very careful engineering of the design, something like this was inevitable.

    • Shirasho@lemmings.world
      link
      fedilink
      arrow-up
      1
      ·
      3 months ago

      I remember when the exact opposite was true. Their drivers were awful and their GPUs consistently overheated.

      I am glad to see things change. Competition is always a good thing.

  • kadu@lemmy.world
    link
    fedilink
    arrow-up
    0
    ·
    edit-2
    3 months ago

    I get the appeal of the “best of the best” but a few years ago I decided to only buy components and tech in general with efficiency in mind, and I’m so happy.

    My RTX 4060 Ti runs everything but stays surprisingly cool for a GPU, gets by with my 500W PSU with power to spare, is stone silent, and everything fits in a nice small form factor case. My computer is silent, cool and wastes very little power. This is also how I’m choosing phones and many other tech gadgets nowadays.

    Having your product be so demanding you need to create a new connector to retrofit into old style power supplies, and then having it melt because even your own adaptor can’t handle the power, is not a good idea at all.

    • huginn@feddit.it
      link
      fedilink
      arrow-up
      1
      ·
      3 months ago

      AFAIK it can handle all the power totally fine - it’s just very easy to partially connect instead of fully.

      • kadu@lemmy.world
        link
        fedilink
        arrow-up
        0
        ·
        3 months ago

        I guess one could make the argument that if it’s so tightly within spec that minor errors can cause catastrophic failure, it can’t really handle it.

        But it can also be said that this is just user error being reported as “Nvidia bad” because this farms clicks and up votes.

        • huginn@feddit.it
          link
          fedilink
          arrow-up
          0
          ·
          3 months ago

          It’s not about how tight the spec is - it’s about how poorly designed the connector is if it is easy to partially connect.

          You can have very tight specs and not be failure prone.

          • kadu@lemmy.world
            link
            fedilink
            arrow-up
            0
            ·
            3 months ago

            If a partial connection, a very common event, is not problematic with other GPUs but very problematic with this one - yes, it’s correct to affirm being so tightly within spec is a problem, as deviations in real world usage are more than expected.

            • huginn@feddit.it
              link
              fedilink
              arrow-up
              1
              ·
              3 months ago

              It’s not that common with previous connectors though. They snapped in and without a full connection did not work.