[FFmpeg-devel] [PATCH] SIMD-optimized float_to_int32_fmul_scalar()

Justin Ruggles justin.ruggles
Fri Jan 7 19:15:37 CET 2011


This patch implements float_to_int32_fmul_scalar() for 3dnow, sse, and
sse2 and uses it in the AC3 encoder.

benchmarks (in dezicycles) for scale_coefficients() in ac3enc_float.c:

AMD Athlon 64 X2 6000+ (64-bit Ubuntu)
    C: 137485
3DNow:  52110
  SSE:  50257
 SSE2:  53306

Intel Atom 330 (64-bit Ubuntu)
    C: 595011
  SSE: 149121
 SSE2: 148662

Thanks,
Justin

-------------- next part --------------
A non-text attachment was scrubbed...
Name: float_to_int32_fmul_scalar.patch
Type: text/x-patch
Size: 6611 bytes
Desc: not available
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20110107/b8447693/attachment.bin>



More information about the ffmpeg-devel mailing list