Windows version has 1/2 the keys/sec as linux version.

[Immrspy]

You've reported 72 million key/sec on 16/32 threads of a 5950X, but using my 7900X and 22/24 threads, I only see ~30 million key/sec. Both GUI and CLI give the same result. The only difference is that I am running on Windows. Is the windows optimization that poor, or is there a configuration that I am missing?

Running in Windows Subsystem for Linux, I get 49 million key/sec, closer to 72 million but I would expect to meet or exceed with my processor.

Obviously Windows optimization doesn't need to be a priority, but it would be a nice touch if you're looking for something to develop on for this project.

[hinto-janai]

IIRC the 5950X measurement was taken using 32 threads, very slightly undervolted (lower temps, allows for higher clock speed boost).

The Windows version may be slower although I believe the thread count / clock speed is making a difference here as well. Maybe CPU cache and memory speed differences too but I'm less sure on that.