Sunday, September 18, 2011 11:16 AMI understand that a primary platform for C++ AMP is a GPU, but in many use cases multicore CPU with SIMD can be an attractive target as well. Does C++ AMP autovectorize to take advantage of SSE or AVX?
- Changed Type DanielMothMicrosoft Employee, Owner Friday, April 06, 2012 7:22 AM
Sunday, September 18, 2011 2:49 PMAccording to the Microsoft guy I spoke with at BUILD, it does indeed. There may be some limitations as to what can be vectorized, and he only mentioned SSE, not SSE2 or AVX.
Tuesday, September 20, 2011 2:50 AMOwner
Please read this post explaining the story on our CPU fallback and please post back here if you have further questions:
- Marked As Answer by DanielMothMicrosoft Employee, Owner Friday, April 06, 2012 7:22 AM