I’m curious as to what is causing different GPU rendering results. I’m running a bunch of tests on various GPUs (3090, 3080s, 2080s, sli, etc.)
1. I assembled a newer computer, my first AMD in over 12 years. It’s an AM4 5900 on a new MB that supports PCIE v4.0. 32gb RAM (newer ram I forget which kind.)
2. My older computer is from 2018 with top of the line Intel hardware. Pcie 3.0 max. 64 gb ram. Still very good hardware.
3. Both the same exact OS and nvidia studio drivers.
4. Older computer with an EVGA 2080 Black TI (12gb onboard) scores around 1200 on the “older computer.” Newer computer the same card hits 1361.
I understand it’s different underlying and faster hardware on the newer machine, but the 2080 supports pcie v3.0 max and I would think that since most of the processing occurs only on the GPU, the speeds should be near identical. Only thing I can think of is that the newer computer is less bottlenecked somewhere? Other than r/w speed to onboard ram and a newer nvme, what could cause this? I didn’t think those made that much of a difference. I have yet to test the higher end GPUs which take advantage of pcie 4.0.
1. I assembled a newer computer, my first AMD in over 12 years. It’s an AM4 5900 on a new MB that supports PCIE v4.0. 32gb RAM (newer ram I forget which kind.)
2. My older computer is from 2018 with top of the line Intel hardware. Pcie 3.0 max. 64 gb ram. Still very good hardware.
3. Both the same exact OS and nvidia studio drivers.
4. Older computer with an EVGA 2080 Black TI (12gb onboard) scores around 1200 on the “older computer.” Newer computer the same card hits 1361.
I understand it’s different underlying and faster hardware on the newer machine, but the 2080 supports pcie v3.0 max and I would think that since most of the processing occurs only on the GPU, the speeds should be near identical. Only thing I can think of is that the newer computer is less bottlenecked somewhere? Other than r/w speed to onboard ram and a newer nvme, what could cause this? I didn’t think those made that much of a difference. I have yet to test the higher end GPUs which take advantage of pcie 4.0.
Comment