0

I own Intel Parallel Studio 2016, but I find the compiler way too slow, unreliable and actually providing worse code than Clang, but the SVML is a great thing, so I'm trying to use it via Agner Fog's VectorClass (https://www.agner.org/optimize/). Things like sine and cosine seems working, but pow __svml_powf4 doesn't. It basically produces a pile of nonsense. __svml_powf8 is the same...

Any luck using these? Could there be a problem with calling convention? (considering it has 2 arguments, while sin/cos have just one)

Vojtěch Melda Meluzín
  • 1,117
  • 3
  • 11
  • 22
  • Can you provide any more information? Which platform is this on? What hardware instruction set are you compiling for? Which compilers are you using, just Clang? What is the error that you are seeing? Basically any further information you can provide will make it more likely you'll get an answer which can help with this. – Antony Peacock Nov 04 '18 at 14:42
  • Windows, AVX/AVX2/AVX512, Clang. Returned values make no sense. Anyways I started using the math functions in VectorClass and it works great so no need for SVML anymore. – Vojtěch Melda Meluzín Nov 05 '18 at 15:06
  • According to the intel Intrinsics Guide, SVML intrinsics should start with `_mm`. Which header are you getting `__svml` intrinsics from? – Dmitri Nesteruk Jun 24 '20 at 17:18

0 Answers0