Blender Git Loki
Git Commits -> Revision ac908f6
Revision ac908f6 by Thomas Dinges (soc-2014-cycles) May 13, 2014, 08:37 (GMT) |
Cycles: Add an AVX2 CPU kernel. New optimized kernel, which requires AVX2 and FMA3 instruction sets. At the moment the speedup is small (~2%) as we only use gcc and clang auto optimization, but we can use dedicated intrinsics for that later. D482 would be a good basis for further improvements. |
Commit Details:
Full Hash: ac908f6c1f6d77790d2645104d4ba9a139937317
Parent Commit: 6d62837
Lines Changed: +176, -0
1 Added Path:
/intern/cycles/kernel/kernel_avx2.cpp (+82, -0) (View)
8 Modified Paths:
/intern/cycles/CMakeLists.txt (+7, -0) (Diff)
/intern/cycles/device/device_cpu.cpp (+52, -0) (Diff)
/intern/cycles/kernel/CMakeLists.txt (+2, -0) (Diff)
/intern/cycles/kernel/kernel.h (+11, -0) (Diff)
/intern/cycles/SConscript (+3, -0) (Diff)
/intern/cycles/util/util_optimization.h (+5, -0) (Diff)
/intern/cycles/util/util_system.cpp (+13, -0) (Diff)
/intern/cycles/util/util_system.h (+1, -0) (Diff)
/intern/cycles/device/device_cpu.cpp (+52, -0) (Diff)
/intern/cycles/kernel/CMakeLists.txt (+2, -0) (Diff)
/intern/cycles/kernel/kernel.h (+11, -0) (Diff)
/intern/cycles/SConscript (+3, -0) (Diff)
/intern/cycles/util/util_optimization.h (+5, -0) (Diff)
/intern/cycles/util/util_system.cpp (+13, -0) (Diff)
/intern/cycles/util/util_system.h (+1, -0) (Diff)
Master Commits
MiikaH:n Sivut a.k.a. MiikaHweb | 2003-2021