Mesa OpenGL Threading Work Sees Much Reduced Memory Footprint For OpenGL Calls
Longtime AMD open-source Mesa developer Marek Olšák after more than one decade working officially for AMD and years before that as an independent open-source contributor going back to the R300g days still has not run out of new performance optimizations to pursue. The most recent accomplishment for this leading Mesa contributor are some refinements to the OpenGL threading "glthread" code for lowering the memory footprint.
Marek noted in a recent and since merged pull request for Mesa 24.1:
The merge request to lower the memory footprint of OpenGL calls in the OpenGL threading mode also has a number of clean-ups and fixes. In total 32 patches were merged on Friday for this nice incremental improvement. This work benefits not only AMD open-source drivers but all Mesa glthread users.
Look for Mesa 24.1 to debut around April with this glthread work and much more across the growing number of OpenGL and Vulkan drivers.
Marek noted in a recent and since merged pull request for Mesa 24.1:
"The biggest change is that the memory footprint of GL calls is decreased substantially. In one viewperf case (which could be the most extreme case), the number of written glthread slots per frame is decreased by 51%, and thus the number of filled batches per frame is decreased by the same number."
The merge request to lower the memory footprint of OpenGL calls in the OpenGL threading mode also has a number of clean-ups and fixes. In total 32 patches were merged on Friday for this nice incremental improvement. This work benefits not only AMD open-source drivers but all Mesa glthread users.
Look for Mesa 24.1 to debut around April with this glthread work and much more across the growing number of OpenGL and Vulkan drivers.
11 Comments