Giters

DLTcollab / sse2neon

A translator from Intel SSE intrinsics to Arm/Aarch64 NEON implementation

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

AVX support?

appwagner opened this issue · comments

commented

Thanks very much for creating this useful project. I'm trying to convert some code that uses AVX instructions using _m256d registers and it doesn't look like this type is handled in the current version of sse2neon. Thanks again for your help.

commented

As the project name implies, SSE2NEON tends to provide SSE coverage rather than AVX. You can use SIMDe for generic AVX support. Also, you can browse the source code of embree to check how AVX code is mapped to Arm NEON counterpart.

commented

Thanks for the references. I'm actually porting this code on an M1 mac so it may be better to use the Apple Accelerate frame work from the start for my use case.