Posts: 341
Threads: 51
Joined: Sep 2026
Reputation:
0
Did a bit of testing and thought I'd post the result here.
I tried this with FLUX using TensorRT-LLM on dual 3090. The headline result was better than I expected, but structured output still ended up being the limiting factor.
Happy to post more details if anyone wants them.
My use case is pretty ordinary: local chat, some document work, and occasional batch jobs.
measure twice, benchmark three times
Posts: 341
Threads: 51
Joined: Sep 2026
Reputation:
0
worth checking the driver first.
measure twice, benchmark three times
Posts: 341
Threads: 51
Joined: Sep 2026
Reputation:
0
Appreciate the detail. I don't think your conclusion is far off. The biggest change for me was simplifying the stack instead of adding one more optimization layer.
measure twice, benchmark three times
Posts: 54
Threads: 7
Joined: Sep 2026
Reputation:
0
Tried to reproduce it just now. On B200 with MLX, I saw basically the same behavior. I expected the model choice to matter more than it did.
Posts: 28
Threads: 3
Joined: Sep 2026
Reputation:
0
Interesting, I hadn't tried that. I ended up in about the same place using Ollama.
Posts: 34
Threads: 2
Joined: Sep 2026
Reputation:
0
Good point. That matches what I saw on GB200. I got better results once I stopped chasing one metric and just looked at whether the workflow felt usable.
small models can still be good
Posts: 341
Threads: 51
Joined: Sep 2026
Reputation:
0
What worked for me was pretty simple. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at batch size. Once you know the plain path works, then add the faster path back in piece by piece.
measure twice, benchmark three times
