Generalize MAC vectorization: either loop order, f16/f32/mixed dtypes, fold into --disable-dsd - #75

Generalize MAC vectorization: either loop order, f16/f32/mixed dtypes, fold into --disable-dsd#75

ycmath wants to merge 6 commits into

spcl:mainspcl/spada:mainfrom

ycmath:nd-mac-vectorizationycmath/spada:nd-mac-vectorizationCopy head branch name to clipboard