Simd
Simd copied to clipboard
armV8 gcc compiler auto-vectorization faster than Neon Intrinsics
I am confused,why gcc auto-vectorization is faster than neon instrinsics. I test open source code like NCNN .