AVX2 implementation for highbd_convolve_2d
Can be up to >10% faster with bit exact results Change-Id: I5f169673fd2d5af96f425f00d862f3c989228d2e
Showing
This diff is collapsed.
Please register or sign in to comment
Can be up to >10% faster with bit exact results Change-Id: I5f169673fd2d5af96f425f00d862f3c989228d2e