Sergey Sharybin c08727ebab Cycles: Quick experiment with using feature-adaptive kernels for CUDA ... Gives few percent of memory improvement for regular feature set kernel and could give significant memory improvement for Experimental kernel. It could also give some degree of performance improvement, but this I didn't really measure reliably yet. Code is ifdef-ed for now, since it's only working on Linux and requires CUDA toolkit to be installed (other platform only use precompiled kernels). This is just an experiment for now and a base for the proper feature support in the future (with runtime compilation using CUDA 7?).		2015-11-21 22:16:01 +05:00
..
CMakeLists.txt	Cycles: Synchronize changes with standalone repository	2015-01-01 01:31:07 +05:00
device_cpu.cpp	Cycles: Prepare for more image extension types support	2015-07-28 14:14:24 +02:00
device_cuda.cpp	Cycles: Quick experiment with using feature-adaptive kernels for CUDA	2015-11-21 22:16:01 +05:00
device_intern.h	Cycles: Repot CPU and CUDA capabilities to system info operator	2015-01-06 14:13:21 +05:00
device_memory.h	Cycles: Gracefully handle out-of-memory happening in device vector	2015-10-11 19:41:59 +05:00
device_multi.cpp	Cycles: Prepare for more image extension types support	2015-07-28 14:14:24 +02:00
device_network.cpp	Cycles: Fix for typo in previous commit	2015-08-10 17:24:14 +02:00
device_network.h	Cleanup: Fix Cycles Apache header.	2014-12-25 02:50:24 +01:00
device_opencl.cpp	Cycles: Move build options constructions to DeviceRequestedFeatures	2015-11-21 21:42:31 +05:00
device_task.cpp	Cycles: Code cleanup, spaces around keywords	2015-03-28 00:15:15 +05:00
device_task.h	Cycles: Pass requested tile size to the device via device task	2015-05-09 19:09:07 +05:00
device.cpp	Cycles: Log requested device features	2015-07-18 16:02:09 +02:00
device.h	Cycles: Move build options constructions to DeviceRequestedFeatures	2015-11-21 21:42:31 +05:00