For the comparison to os x docker, the biggest slowdown for os x docker is the file system mapping. It basically breaks any reasonable performance chart: look up posts getting into alternatives like mutagen. The slowdown makes using an IO-heavy dev flow like live editing via `npm run watch` relatively unusable: 30X+ slowdowns are normal. There might be high CPU overheads too (I recall we had to tweak seccomp), but when basic IO is so slow, the rest doesn't matter.

In contrast, yes, wsl2 has 5% CPU hits (hyperv, ...),but a sane FS mapping, so the total overhead is imperceptible for a windows dev box.

I was pleasantly surprised to see wsl2 Just Work. Our only issue preventing wsl2 from being the official team rec over native Linux has been wsl2's lack of opencl, and that's just specific to our use of GPUs. As someone whose preferred dev box has been osx for ~20 years, even when at MS, I was biased against Windows for most dev... but no longer.

HyperV is lightweight and from what I understood, windows runs on top of hyperV too when you enable WSL2. You may lose a few % of performance in video games.

In theory yes, and we got tiny programs running this way, but could not afford to keep subsidizing megacorp politics.

Nvidia punts to IBM RHEL8 docs for GPU podman, which is unusual and risky to see. We officially recommend against it for HA environments due to this kind of lack and overall low relative confidence. I think k8s envs may be moving to something here, so maybe in a year or two? I'd be curious of folks doing stock rhel8 podman with tensorflow/torch on nvidia, which should be as vanilla as you can get for enterprise ai. We generally see more interesting GPU envs here (ex: DGX with advanced networking hw/sw), but we don't have confidence for the simple case, which is the starting point..