Agner's CPU blog

I tried replacing all "movdqa reg1, reg2" with "pshufd $0xE4, reg1, reg2" and I'm still getting same results. I also tried to see if forced context switching would allow performance to recover. I modified program to do the YMM instruction + VZEROUPPER just at the beginning of the program, before tes...

I've ran into interesting behaviour on Zen4 (7900X) with mixed VEX and non-VEX code. It seems that, in 32-bit mode, VZEROUPPER is not fully clearing some state for YMM/XMM registers which causes non-VEX code run at slower speed after YMM usage. I'm running following code in 32-bit mode and 64-bit mo...

agner wrote: ↑

2020-05-23, 5:11:26

I will be happy to test it if somebody gives me access to it. I can test it through a remote login to a Linux machine.

I can provide remote access to a Linux machine with Ryzen 3700X. Send me PM.