[svn] Repository:
ViewVC logotype

Revision 1205


Jump to revision: Previous Next
Author: edgomez
Date: Thu Nov 13 23:11:24 2003 UTC (15 years, 1 month ago)
Log Message:
MMXed the calculation of SSE for 8x8 16bit blocks. This helps quite
a lot VHQ=4 mode.

My tests show with trellis:chroma_me:
  - ~20% speed improvement for vhq=4.
  - at least 5% when using vhq=1.

Of course this speedup vanishes if more CPU intensive features are
used. CruNcher who used gmc/qpel, noticed "only" a ~5% speed
improvement.

NB: i'm of course talking about overall speed improvement. Such a
small patch for such a big improvement :-)

Changed paths:

Path Details
Directorybranches/dev-api-4/xvidcore/src/motion/estimation_rd_based.c modified , text changed
Directorybranches/dev-api-4/xvidcore/src/motion/sad.c modified , text changed
Directorybranches/dev-api-4/xvidcore/src/motion/sad.h modified , text changed
Directorybranches/dev-api-4/xvidcore/src/motion/x86_asm/sad_mmx.asm modified , text changed
Directorybranches/dev-api-4/xvidcore/src/xvid.c modified , text changed

No admin address has been configured
ViewVC Help
Powered by ViewVC 1.0.4