Hi, guys,
My previous system consisted of a CPU with 16 lanes, an MSI Golden GTX 980 Ti, a GTX 650 PhysX slave, and an M.2 x4 SSD. The M.2 requires 4 of 16 CPU lanes, which means that the only way the GPUs could work is by running in x4 mode instead of x8 mode, and I think that was the cause of the bottleneck.
1. Would running those GPUs in x4 instead of x8 cause a considerable bottleneck?
2. Would running the PhysX card on a PCIe 2.0 x16 lane instead of a 3.0 lane drop its lanes down to 4 so the 980 Ti can utilize a total of 8, or would I literally need a x16 to x4 riser card to force the PhysX card to run in x4?
3. Assuming the PSU is adequate would having a PhysX card in x4 hurt the system more than simply removing the card altogether?
My previous system consisted of a CPU with 16 lanes, an MSI Golden GTX 980 Ti, a GTX 650 PhysX slave, and an M.2 x4 SSD. The M.2 requires 4 of 16 CPU lanes, which means that the only way the GPUs could work is by running in x4 mode instead of x8 mode, and I think that was the cause of the bottleneck.
1. Would running those GPUs in x4 instead of x8 cause a considerable bottleneck?
2. Would running the PhysX card on a PCIe 2.0 x16 lane instead of a 3.0 lane drop its lanes down to 4 so the 980 Ti can utilize a total of 8, or would I literally need a x16 to x4 riser card to force the PhysX card to run in x4?
3. Assuming the PSU is adequate would having a PhysX card in x4 hurt the system more than simply removing the card altogether?