My favorite is that AVX2's ALIGNR is very different from SSE's ALIGNR. https://github.com/rust-lang/regex/blob/60d087a23025e045ae754a345b04003c31d83d93/src/vector/avx2.rs#L164-L166 …
-
-
This Tweet is unavailable.
-
- 1 more reply
-
-
-
Why? Because 32*32 multiplication has 64 bit results? Or just because reasons...
-
Yes, you would need a __mm256i return value which was not really a thing in pre-AVX times. I think that interface trade-off when switching from addition to multiplication is often seen in low-level APIs
End of conversation
New conversation -
-
-
Considering it’s matrix-matrix wouldn’t the result be signed integers that are 64 bit
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.