2012/1/12 Christophe Gisquet <christophe.gisq...@gmail.com>: > 2012/1/12 Kostya Shishkov <kostya.shish...@gmail.com>: >> Probably OK, I'll wait for other people to say something. Also is there a >> chance of doing benchmarks? > > Before: 4.134 / 4.618 / 2.668 / 1.123 / 3.120 > After: 3.916 / 4.399 / 2.558 / 1.092 / 2.995 > So around a 4% speed increase.
For core i5@2.53. For core2@3.16 and other sequences: MD5=1b14ff1ac5163365f0bcc3af26f53729 3.236 3.180 MD5=336a974a043d246b97cdb99ec89040d9 3.572 3.472 MD5=e04cdb4e55c8715e84ccdab99ccc81ba 0.900 0.876 MD5=4815fbdd11b34c49ecb93c41f421910e 4.516 4.456 MD5=51ac62e25c188f8c4344f37b8ebf95a9 2.100 2.068 MD5=f5d031f5253e6c97e5d6e6a44794b9b0 2.484 2.456 So here, more like 2% gain. Christophe _______________________________________________ libav-devel mailing list libav-devel@libav.org https://lists.libav.org/mailman/listinfo/libav-devel
