It’s likely wrong reporting, the m5 ultra gets 614 GB/s today. The m3 pro gets over 800. The rumored m5 pro is 1.2TB/s. Likely this will be 2.4 TB/s, but half the reason for high bandwidth in nvidia chips is somewhat offset in the unified scheme. Mlx needs a lot less copy data around when the gpu can just read it directly.

/dev/null compresses everything perfectly. The hard part is recovery and is left to the reader.

This page intentionally left blank.

Doesn’t do much if the process is in uninterruptible I/o. Kill can’t do much against syscalls that aren’t returning, you’ll be rebooting. You can try this https://github.com/mitchty/nix/blob/shenanigans/src/evilkill.sh but it’s not guaranteed to work. It’s just trying to force the process to exit(), but it doesn’t always work and can’t. All that script amounts to is attaching a debugger and executing assembly to mimic exit. I’ll say 60% of the time it works 100% of the time when kill doesn’t.

Sigh, you missed the obvious pun: would you like some cheese with your wine?

As a nixos user I guess so. It is like -5 here though so today it’s more heated blanket.

I use shenanigans, more fitting and descriptive.

I’m using shenanigans now, fits the best methinks.