Hi, I wrote this code before we found new algorithm. Sure, I will compare with new algorithm once I finish remaining modes which don't have AVX2 asm.
On Tue, Aug 18, 2015 at 8:50 PM, chen <chenm...@163.com> wrote: > This is old algorithm, it need transpose, could you compare to new > algorithm? > > > At 2015-08-18 12:11:35,dnyanesh...@multicorewareinc.com wrote: > ># HG changeset patch > ># User Dnyaneshwar G <dnyanesh...@multicorewareinc.com> > ># Date 1439531917 -19800 > ># Fri Aug 14 11:28:37 2015 +0530 > ># Node ID 5ed23f786ea8f98e003189a537f960e4ff16201f > ># Parent 996ebce8c874fc511d495cee227d24413e99d0c1 > >asm: AVX2 asm for intra_ang_32 mode 9, improved over 40% than SSE asm > > > >updated intra_ang_32 mode 27 AVX2 asm code, improved over 3% than previous > >AVX2 code > >removed unnecessary constants from previous asm > > > _______________________________________________ > x265-devel mailing list > x265-devel@videolan.org > https://mailman.videolan.org/listinfo/x265-devel > >
_______________________________________________ x265-devel mailing list x265-devel@videolan.org https://mailman.videolan.org/listinfo/x265-devel