Hey all, this morning I've found 5 minutes to run a quick test on one of our more complex archviz scenes and compare CPU and GPU performance.
The scene takes 20GB of RAM.
I run a progressive render with following settings.
msr: 24
max subdivs: 70
noise treshold set to 0.009 (probably too low but I don't it matters in this case)
max render time of 4 minutes.
Ray bundle size: 128 on CPU and 192 / 16 on GPU - I left those at default.
V-RAY CPU: 2 x i9 7900X
V-RAY GPU RTX: 2 x 2060 SUPER + 3970X (I am not sure if it was utilized in RTX mode)
I started renders simultaneously and as expected CPU started frame faster, although only maybe 20 seconds faster then GPU. Light cache calculation was about 20% faster on CPU but this I think it is related to i9 7900X handling it better than 3970X (please correct me if I am wrong).
As you can see from the images the GPU is significantly cleaner.
On LITTLE thing...on the second frame of animation GPU stopped with an Optix error...
I also uploaded the second frame to give you an idea of what kind of Archviz environment it is.
Comment