Coefficient decoding/msac optimizations
Around 1% faster overall on Skylake, though modern Intel CPU:s have really good unaligned memory performance so the buffer alignment might potentially be more beneficial on some older CPU:s.
Around 1% faster overall on Skylake, though modern Intel CPU:s have really good unaligned memory performance so the buffer alignment might potentially be more beneficial on some older CPU:s.
VideoLAN code repository instance