+- Inference (https://www.datasculpting.ai/inference)
+-- Forum: Inference Engines (https://www.datasculpting.ai/inference/forumdisplay.php?fid=9)
+--- Forum: vLLM / SGLang (https://www.datasculpting.ai/inference/forumdisplay.php?fid=36)
+--- Thread: vLLM vs SGLang for small teams on dual 3090 (/showthread.php?tid=1352)
Pages:
1
2
vLLM vs SGLang for small teams on dual 3090 - phasespark - 10-16-2026
This keeps coming up in conversation so I figured I'd start a thread.
One thing I keep noticing is that people underestimate long context until they actually try to run this stuff for more than one afternoon.
Curious whether that lines up with what other people are seeing.
Re: vLLM vs SGLang for small teams on dual 3090 - batchsize - 10-16-2026
I'd start with the simpler stack.
Re: vLLM vs SGLang for small teams on dual 3090 - ggufvault - 10-17-2026
I tested this tonight. I couldn't reproduce the slowdown until I pushed context much higher. The surprise was that memory pressure showed up before throughput collapsed.
Re: vLLM vs SGLang for small teams on dual 3090 - cacheline - 10-17-2026
That matches what I was seeing. The pattern around NVFP4 has been pretty similar for me. Worth rechecking your defaults though, because one bad assumption can make the rest of the thread look confusing.
Re: vLLM vs SGLang for small teams on dual 3090 - tinycontext - 10-17-2026
A couple things I'd check first. If you're on SGLang, I'd confirm the version first and then reduce variables until the behavior is obvious. Not glamorous advice, but repeatable baselines save a lot of time.
There are a few different conversations hiding inside this thread, so being explicit about the goal helps.
Re: vLLM vs SGLang for small teams on dual 3090 - fusedhall56 - 10-17-2026
Not sure I agree there. The reported numbers can look compelling, but in ordinary usage I didn't see the same benefit. Worth testing the simpler explanation before adopting the more interesting one.
Re: vLLM vs SGLang for small teams on dual 3090 - batchsize - 10-17-2026
A couple things I'd check first. If you're on Ollama, I'd confirm the version first and then reduce variables until the behavior is obvious. Not glamorous advice, but repeatable baselines save a lot of time.
Re: vLLM vs SGLang for small teams on dual 3090 - fastbench - 10-17-2026
Appreciate the detail. 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: vLLM vs SGLang for small teams on dual 3090 - batchsize - 10-18-2026
Quote:Appreciate the detail. I don't think your conclusion is far off. Worth rechecking your defaults though, because one bad assumption can make the rest o
worth checking the driver first.
If you post your exact settings, people here can probably narrow it down pretty fast.
Re: vLLM vs SGLang for small teams on dual 3090 - thetashed24 - 10-18-2026
My first guess would be this: Before rebuilding anything, I'd test one clean baseline with DeepSeek and compare from there. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.