bartvdbraak/blender
1
0
Fork 4
Code Issues 4 Pull Requests 1 Actions 1 Packages Projects Releases Wiki Activity

Fix OpenCL group size performance issue on Intel GPUs

Browse Source 
Contributed by Intel. On some scenes like classroom with particular integrated
GPUs this speeds up rendering 1.97x. With other benchmarks and GPUs it's
between 0.99-1.14x.
This commit is contained in:
Brecht Van Lommel 2021-05-17 13:32:50 +02:00
parent bb32ecadb5
commit 91a5dbbd17
1 changed files with 5 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
intern/cycles/device/opencl/device_opencl_impl.cpp
View File

@ -569,6 +569,11 @@ class OpenCLSplitKernel : public DeviceSplitKernel {
size_t num_elements = max_elements_for_max_buffer_size(kg, data, max_buffer_size);
int2 global_size = make_int2(max(round_down((int)sqrt(num_elements), 64), 64),
(int)sqrt(num_elements));
if (device->info.description.find("Intel") != string::npos) {
global_size = make_int2(min(512, global_size.x), min(512, global_size.y));
}
VLOG(1) << "Global size: " << global_size << ".";
return global_size;
}