I wonder how well branch predictors handle CLZ vs just masking and comparing bytes. I expect the latter to work really well, but who knows how optimized CLZ-

I wonder how well branch predictors handle CLZ vs just masking and comparing bytes. I expect the latter to work really well, but who knows how optimized CLZ-derived branches are.