xenonite
I've made a few optimized compiles of the latest Open-Source svpflow1.dll library with the 2015 version of the Intel Compiler on Visual Studio 2013
You can just share your settings cause we're using exactly the same toolchain ![]()
And I don't expect "AVX" could be any faster than "SSE2" one, still the main bottleneck is in x264 asm code.
I only post them incase there are some users that would like an (at least partially) ICL compiled SVP
SVPflow libs are compliled with ICL since ver.1.1.14 - http://forum.doom9.org/showthread.php?p … ost1718861