When doing the AVX2 implementation, min/max across a vector will come in handy. Here is a SIMD horizontal min or max: http://stackoverflow.com/questions/23590610/find-index-of-maximum-element-in-x86-simd-vector
When doing the AVX2 implementation, min/max across a vector will come in handy. Here is a SIMD horizontal min or max: http://stackoverflow.com/questions/23590610/find-index-of-maximum-element-in-x86-simd-vector