diff options
author | Stefan Werner <stefan.werner@tangent-animation.com> | 2019-03-14 13:45:58 +0300 |
---|---|---|
committer | Stefan Werner <stefan.werner@tangent-animation.com> | 2019-03-14 13:45:58 +0300 |
commit | 47da8dcbcad4ccc5349bc303394e1d01d1c822c5 (patch) | |
tree | ab41c5c2b20e52607e187e4565132d3a5fb4255c | |
parent | 4887baf7d69c6c44c265044600ec00c77e9aa406 (diff) |
Cycles: Improved thread order for better CUDA performance.
This patch puts threads that render the same pixel closer together,
as opposed to threads that render the same sample. Thus threads
within a warp are more coherent in memory access and control flow,
leading to performance improvements.
Example benchmarks on a Quadro RTX4000 (WDDM) on Windows 10:
Koro: 4:23 -> 3:46
BMW: 1:18 -> 1:25
Barbershop Interior: 17:52 -> 14:55
Classroom: 4:37 -> 3:45
Performance differences on OpenCL/AMD were hit and miss, some scenes
became faster, others lost significantly. Therefore, this is kept as
CUDA only change for now.
-rw-r--r-- | intern/cycles/kernel/kernel_work_stealing.h | 6 |
1 files changed, 6 insertions, 0 deletions
diff --git a/intern/cycles/kernel/kernel_work_stealing.h b/intern/cycles/kernel/kernel_work_stealing.h index 0c2d9379b63..9667156eaf5 100644 --- a/intern/cycles/kernel/kernel_work_stealing.h +++ b/intern/cycles/kernel/kernel_work_stealing.h @@ -66,9 +66,15 @@ ccl_device_inline void get_work_pixel(ccl_global const WorkTile *tile, ccl_private uint *y, ccl_private uint *sample) { +#ifdef __KERNEL_CUDA__ + /* Keeping threads for the same pixel together improves performance on CUDA. */ + uint sample_offset = global_work_index % tile->num_samples; + uint pixel_offset = global_work_index / tile->num_samples; +#else /* __KERNEL_CUDA__ */ uint tile_pixels = tile->w * tile->h; uint sample_offset = global_work_index / tile_pixels; uint pixel_offset = global_work_index - sample_offset * tile_pixels; +#endif /* __KERNEL_CUDA__ */ uint y_offset = pixel_offset / tile->w; uint x_offset = pixel_offset - y_offset * tile->w; |