2 ms·
Does this break vectorization? Specifically, saxpy (Y=M*X+B) requires extra operations to dereference every element and won't let you insert an FMA instruction.
by frozenport 13y ago
Does this break vectorization? Specifically, saxpy (Y=M*X+B) requires extra operations to dereference every element and won't let you insert an FMA instruction.