
Nvidia wins unless some other company starts selling cheaper, faster, more efficient matrix multiplication machines.
I’ve read some articles about radically different inference architectures that may tilt the scales, but I know this is wishful thinking because I really would like Nvidia to fail badly.
Linus_nvidia.gif



Sure, I’m going to start running my containers using WSL. Who knows what kind of wonderful surprises are in store in the network layer? How the environment parameter passing will be somewhat more brittle with funky new ways to break because of new characters you have to escape? I can’t wait to troubleshoot all of that, and spend countless hours on MS forums and meet fellow colleagues with similar issues. What fun.