+- Inference (https://www.datasculpting.ai/inference)
+-- Forum: Optimization & Performance (https://www.datasculpting.ai/inference/forumdisplay.php?fid=10)
+--- Forum: Context / KV Cache (https://www.datasculpting.ai/inference/forumdisplay.php?fid=41)
+--- Thread: speculative decoding actually worth it on dual 3090 (/showthread.php?tid=1085)
speculative decoding actually worth it on dual 3090 - cacheline - 10-11-2026
Did a bit of testing and thought I'd post the result here.
After a few runs on GB200, I ended up with decent performance, but the setup only really made sense once I adjusted around reranking.
Curious if anyone else landed in the same place.
My use case is pretty ordinary: local chat, some document work, and occasional batch jobs.
Re: speculative decoding actually worth it on dual 3090 - quantnest - 10-11-2026
What exact setup are you running? What model, what runtime, and roughly what context size?
Re: speculative decoding actually worth it on dual 3090 - cacheline - 10-12-2026
Good point. 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: speculative decoding actually worth it on dual 3090 - cacheline - 10-12-2026
Quote:Good point. The pattern around NVFP4 has been pretty similar for me. The biggest change for me was simplifying the stack instead of adding one more op
What worked for me was pretty simple. On dual 3090, the easiest win is often to back off the aggressive settings and see whether the problem survives. Not glamorous advice, but repeatable baselines save a lot of time.
Re: speculative decoding actually worth it on dual 3090 - nanopatch - 10-13-2026
Not sure I agree there. I wouldn't optimize around that first. On my side the bigger swing came from the runtime and not the model.
Re: speculative decoding actually worth it on dual 3090 - cacheplane98 - 10-13-2026
A couple questions before I guess too hard. Are you testing this on 5090 or something else?
Re: speculative decoding actually worth it on dual 3090 - cacheline - 10-13-2026
Quote:A couple questions before I guess too hard. Are you testing this on 5090 or something else?
Tried to reproduce it just now. I couldn't reproduce the slowdown until I pushed context much higher. I expected the model choice to matter more than it did.
Re: speculative decoding actually worth it on dual 3090 - dockbrook21 - 10-15-2026
My first guess would be this: Before rebuilding anything, I'd test one clean baseline with Mixtral and compare from there. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.
Re: speculative decoding actually worth it on dual 3090 - oxidestream - 10-15-2026
A couple things I'd check first. Before rebuilding anything, I'd test one clean baseline with Llama and compare from there.
Re: speculative decoding actually worth it on dual 3090 - cacheline - 10-15-2026
A couple things I'd check first. On GB200, the easiest win is often to back off the aggressive settings and see whether the problem survives. Once you know the plain path works, then add the faster path back in piece by piece.