+- 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: early impressions (/showthread.php?tid=557)
Pages:
1
2
Re: NVIDIA driver regression: early impressions - sparrowkite53 - 10-03-2026
Quote:subscribed, curious where this lands.
I'd approach it in a boring order. If you're on MLX, I'd confirm the version first and then reduce variables until the behavior is obvious.
Re: NVIDIA driver regression: early impressions - kernelmike - 10-03-2026
Quote:Quote:subscribed, curious where this lands.
Thanks, that helps. I don't think your conclusion is far off. Worth rechecking your defaults though, because one bad assumption can make the rest of the thread look confusing.
Re: NVIDIA driver regression: early impressions - cobaltspark - 10-03-2026
Quote:Quote:Quote:subscribed, curious where this lands.
That matches what I was seeing. The pattern around NVFP4 has been pretty similar for me. The biggest change for me was simplifying the stack instead of adding one more optimization layer.
Re: NVIDIA driver regression: early impressions - audiopath66 - 10-03-2026
Quote:Quote:Quote:Quote:subscribed, curious where this lands.
Maybe, but I had the opposite result. I've seen people blame tool calling when the real issue was much more mundane. Worth testing the simpler explanation before adopting the more interesting one.
Re: NVIDIA driver regression: early impressions - dockember - 10-03-2026
following this because I'm in the same boat.
Re: NVIDIA driver regression: early impressions - audiopath66 - 10-03-2026
Ran a quick check on my box. My result was mixed: first run looked bad, second run looked normal. The surprise was that memory pressure showed up before throughput collapsed.