I have two rigs running the same model with only slightly different settings. Both are i9 CPUs, exact same MB's, etc. The only difference is one has an RTX 3080 10GB and one has an RTX 3060 12GB.
Both units shows GPU VRAM utilized to about 80% roughly.
The unit with the 3080 shows little CPU usage but shows the GPU 3D running 50-60%
The unit with the 3060 shows CPU usage at 40-50% and no GPU 3D usage
I would have expected the 3060, with less CUDA cores, to have higher GPU usage, not zero. The only other difference I can think of, between 6 and 8 series, is nVidia Physx support. The 3080 runs the same model at a substantially higher speed, obviously. But I'm just curious why the difference in how the load is processed. I fully expected the 3060 GPU/3D to be maxed out.