Brecht Van Lommel e26eb9c93b Cycles: reduce CUDA stack memory access for Maxwell and up, increasing max registers.
For non-branched path tracing with a GTX 960 and CUDA 7.5, this gives a small reduction
in stack usage but mainly: 8% faster render on BMW, 5% on pabellon, 13% on classroom.
2016-06-19 20:17:26 +02:00
2016-06-17 01:56:59 +10:00
2013-12-24 22:57:27 +06:00
2016-01-04 14:20:48 +05:00
2016-05-21 05:48:10 +10:00
2016-02-10 00:26:12 +11:00
Description
No description provided
841 MiB
Languages
C++ 78%
Python 14.9%
C 2.9%
GLSL 1.9%
CMake 1.2%
Other 0.9%