18-Way NVIDIA GPU Performance With Blender 2.90 Using OptiX + CUDA

A few days ago I published a deep dive into the CPU and GPU performance with Blender 2.90 as a major update to this open-source 3D modeling software. Following that I kept on testing more and older NVIDIA GPUs with the CUDA and OptiX back-end targets to now have an 18-way comparison from Maxwell to Turing with the new Blender 2.90.

Given the Blender 2.90 performance changes over Blender 2.8x as outlined in the earlier article, here is a fresh look at how the NVIDIA GPU performance compares for this large range of graphics processors. Additionally, with Blender 2.90 is now OptiX support for non-RTX GPUs. While RTX GPUs still perform the best with the Blender OptiX support, non-RTX GPUs can now work for this back-end and in some cases perform better than the CUDA back-end.

Tests were done using the Maxwell/Pascal/Turing GPUs I had easily accessible for testing. No word yet on Ampere cards for Linux testing but hopefully we'll have some soon. Though Blender 2.90 does lack support for the RTX 3000 series but hopefully with an updated CUDA build that the next Blender release will have support for these brand new GPUs.




Continue to OpenBenchmarking.org to see the ten benchmarks ran on Blender 2.90 in full across the 18 graphics cards available for testing. From there on OpenBenchmarking.org is also the ability to trivially create performance-per-dollar graphs based on your local pricing or used pricing for the older models, etc.