Performance dilemma

[Dagon]

Option A
- PCIE4 M.2 NVME in slot connected to CPU
- GPU (4090) runs at x8

Option B
- PCIE4 M.2 NVME in slot connected to chipset
- GPU (4090) runs at x16

Which would give the best performance? I know dropping the GPU to x8 loses 1-2% but not sure about difference between CPU/chipset for the M.2

 

[Tetras]

Storage is unlikely to be the bottleneck in the vast majority of use cases, so I would choose option B, even though I doubt you'd notice the difference between either.

Temps might be a consideration between the two slots.

 

[Dagon]

Yeah I think Option B is probably the way to go, thanks.