Hello,
As mentioned in the title, my GPU (2080 TI) is locked at 8x PCI-E lanes, and I don't know exactly why. I have a theory but I need confirmation.
My build :
Mobo : Asrock x570 Taichi
CPU : 3900x
GPU : MSI 2080 TI Gaming X Trio
SSD : Corsair Force MP600 PCI Express 4.0 x4 M.2
RAM : 32GB G.Skill 3600 Mhz CL16
The GPU is installed in the top PCI-E slot and I have no other PCI-E cards. (this is supposed to use 16 CPU PCIe lanes)
The SSD that uses 4 lanes is mounted in the top M.2 slot (straight to CPU, uses 4 lanes).
I also mention that I know the CPU has 24 lanes (16 for GPU, 4 for Nvme and 4 for chipset).
While using multiple utility tools to check the number of lanes used by GPU, they all post "@x8 current usage". I used GPU-Z, Cpu-Z, Speccy and HWInfo.
My theory :
While using HWinfo, the tool shows the pcie slot max bandwidth and the current status. For example : Pci-e x16 8GT/s @ x8 8GT/s. Thing is that my motherboard has PCI-e 4.0 support, while the GPU only knows Pci-e 3.0. I know Pci-e 4.0 is backward compatible with 3.0, and as far as I know, x8 Pcie 4.0 = x16 Pcie 3.0 regarding bandwidth.
In my testing, I noticed that the max bandwidth for PCI-E 3.0 x16 is 8GT/s (is this correct?). My GPU, even though the software shows that it runs at x8 has the same bandwidth as PCI-E 3.0 x16 : 8GT/s.
Can anyone explain this ? Why is this happening ? Is x8 = x16 in my case ?
Thank you
As mentioned in the title, my GPU (2080 TI) is locked at 8x PCI-E lanes, and I don't know exactly why. I have a theory but I need confirmation.
My build :
Mobo : Asrock x570 Taichi
CPU : 3900x
GPU : MSI 2080 TI Gaming X Trio
SSD : Corsair Force MP600 PCI Express 4.0 x4 M.2
RAM : 32GB G.Skill 3600 Mhz CL16
The GPU is installed in the top PCI-E slot and I have no other PCI-E cards. (this is supposed to use 16 CPU PCIe lanes)
The SSD that uses 4 lanes is mounted in the top M.2 slot (straight to CPU, uses 4 lanes).
I also mention that I know the CPU has 24 lanes (16 for GPU, 4 for Nvme and 4 for chipset).
While using multiple utility tools to check the number of lanes used by GPU, they all post "@x8 current usage". I used GPU-Z, Cpu-Z, Speccy and HWInfo.
My theory :
While using HWinfo, the tool shows the pcie slot max bandwidth and the current status. For example : Pci-e x16 8GT/s @ x8 8GT/s. Thing is that my motherboard has PCI-e 4.0 support, while the GPU only knows Pci-e 3.0. I know Pci-e 4.0 is backward compatible with 3.0, and as far as I know, x8 Pcie 4.0 = x16 Pcie 3.0 regarding bandwidth.
In my testing, I noticed that the max bandwidth for PCI-E 3.0 x16 is 8GT/s (is this correct?). My GPU, even though the software shows that it runs at x8 has the same bandwidth as PCI-E 3.0 x16 : 8GT/s.
Can anyone explain this ? Why is this happening ? Is x8 = x16 in my case ?
Thank you