+- Inference (https://www.datasculpting.ai/inference)
+-- Forum: Hardware (https://www.datasculpting.ai/inference/forumdisplay.php?fid=4)
+--- Forum: GPUs (https://www.datasculpting.ai/inference/forumdisplay.php?fid=28)
+--- Thread: PCIe lane weirdness with inference loads (/showthread.php?tid=482)
Pages:
1
2
Re: PCIe lane weirdness with inference loads - turbogrid11 - 09-27-2026
Good point. That matches what I saw on A6000.
Re: PCIe lane weirdness with inference loads - labline - 09-28-2026
What worked for me was pretty simple. Before rebuilding anything, I'd test one clean baseline with Mistral and compare from there. Once you know the plain path works, then add the faster path back in piece by piece.
Re: PCIe lane weirdness with inference loads - cobaltlab - 09-28-2026
Thanks, that helps. That matches what I saw on 4090.
Re: PCIe lane weirdness with inference loads - northhall88 - 09-28-2026
I'd start with the simpler stack.