AVX seems to be consistently slower than SSE4 on AMD parts anyhow, outside of very few special-niche code extremes which synthetically maximize AVX advantages. AVX has been very broken and buggy and patchy since at least the Haswell-E era, as it stands it's not really all that it should be or anything at all like what Intel says it should be.
I think mostly because AVX (especially AVX-512) requires a lot of fine tuning and optimizations to obtain full performance benefits. Fine tuning performance optimizations which, incidentally, Intel claims will be implemented by the non-Intel hardware OEMs and software devs.
"All opinions are not equal. Some are a very great deal more robust, sophisticated and well supported in logic and argument than others." - Douglas Adams
[/Korth]