+- 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 with SGLang (/showthread.php?tid=950)
Pages:
1
2
NVIDIA driver regression with SGLang - dockember - 10-09-2026
Posting this in case it saves someone else a little time.
I'm not trying to build a perfect architecture diagram. I just want a sane starting point around PyTorch and B200.
I'm especially interested in mistakes worth avoiding.
Re: NVIDIA driver regression with SGLang - dockember - 10-10-2026
Appreciate the detail. The pattern around embeddings has been pretty similar for me.
Re: NVIDIA driver regression with SGLang - kernelmike - 10-10-2026
Quote:Appreciate the detail. The pattern around embeddings has been pretty similar for me.
that fixed it for me too.
Re: NVIDIA driver regression with SGLang - oxidemixer - 10-10-2026
That matches what I was seeing. The pattern around prefix caching 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 with SGLang - mintforge - 10-10-2026
that fixed it for me too.
Re: NVIDIA driver regression with SGLang - kernelmike - 10-10-2026
Quote:that fixed it for me too.
Good point. The pattern around long context 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 with SGLang - sparrowkite53 - 10-10-2026
I hit almost the same issue last night.
Re: NVIDIA driver regression with SGLang - thetapath - 10-10-2026
worth checking the driver first.
Re: NVIDIA driver regression with SGLang - pixelspark - 10-11-2026
I hit almost the same issue last night.
Re: NVIDIA driver regression with SGLang - thetapath - 10-12-2026
My first guess would be this: If you're on PyTorch, I'd confirm the version first and then reduce variables until the behavior is obvious.