https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83008
--- Comment #11 from sergey.shalnov at intel dot com --- Richard, “Is this about the "stupid" attempt to use as little AVX512 as possible” No, it is not. I provided asm listing at the beginning with zmm only to illustrate the issue more crystalized. I used "-mprefer-vector-width=512" option to get that asm output. I believe that if we remove these registers (any. Including ymm and xmm) from BB you mentioned - it will be profitable for all architectures. " Sergey, can you test this?" Yes. Going to try this. Sergey