+- 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
Re: NVIDIA driver regression with SGLang - mintforge - 10-12-2026
What worked for me was pretty simple. Before rebuilding anything, I'd test one clean baseline with FLUX and compare from there.
Re: NVIDIA driver regression with SGLang - phaseport36 - 10-12-2026
I tested this tonight. The issue went away once I changed one setting and restarted the stack cleanly. The surprise was that memory pressure showed up before throughput collapsed.
Re: NVIDIA driver regression with SGLang - edgecoast - 10-12-2026
Interesting, I hadn't tried that. That matches what I saw on mini PC. Worth rechecking your defaults though, because one bad assumption can make the rest of the thread look confusing.
Re: NVIDIA driver regression with SGLang - cacheplane98 - 10-12-2026
Ran a quick check on my box. My result was mixed: first run looked bad, second run looked normal.
Re: NVIDIA driver regression with SGLang - mintforge - 10-12-2026
I'd start with the simpler stack.