GCC Intrinsics for Knights Corner vector instructions

GCC Intrinsics for Knights Corner vector instructions


We are an ISV for DNA sequence alignment programs and our current programs use SSE via intrinsics for all dynamic programming. We are wondering when/if the new instructions and registers will be supported in gcc with intrinsics?


3 posts / 0 new
Last post
For more complete information about compiler optimizations, see our Optimization Notice.

The vector intrinsics for the first generation of the Intel Xeon Phi coprocessor will not be incorporated into gcc.

SSE intrinsics may be supported by compilers for avx512 target, but they would give up significant potential performance due to using only 25% of simd width. Recoding with avx512 intrinsics would seem generally less productive than optimizing C.

Leave a Comment

Please sign in to add a comment. Not a member? Join today