+- Inference (https://www.datasculpting.ai/inference)
+-- Forum: Drivers / Frameworks / Toolkits (https://www.datasculpting.ai/inference/forumdisplay.php?fid=8)
+--- Forum: NVIDIA / CUDA (https://www.datasculpting.ai/inference/forumdisplay.php?fid=31)
+--- Thread: NVIDIA driver regression for Phi (/showthread.php?tid=1287)
Pages:
1
2
Re: NVIDIA driver regression for Phi - kernelmike - 10-19-2026
Quote:I tested this tonight. On B200 with MLX, I saw basically the same behavior.
What exact setup are you running? What model, what runtime, and roughly what context size?
Re: NVIDIA driver regression for Phi - phaseport36 - 10-19-2026
A couple things I'd check first. Before rebuilding anything, I'd test one clean baseline with Phi and compare from there. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.