Posting this in case it saves someone else a little time.

If you were starting from scratch today for this use case, what would you choose first and what would you ignore?

Would rather learn from your scars than collect my own all at once.

Did a bit of testing and thought I'd post the result here.

I tried this with DeepSeek using Ollama on 4090. The headline result was better than I expected, but INT4 still ended up being the limiting factor.

Happy to post more details if anyone wants them.

This is mostly for local work rather than public serving.

Spent some time on this today and wanted to compare notes.

I'm comparing two approaches here: keep it simple on Threadripper box, or squeeze more out of the stack and deal with the complexity.

Trying to avoid rebuilding the whole thing twice.

I care more about stability than squeezing the last possible tokens/sec out of it.

Did a bit of testing and thought I'd post the result here.

Short version: llama.cpp behaved well for me, but only after I stopped ignoring router policy.

Curious if anyone else landed in the same place.

Wanted to compare notes with people doing similar things.

I'm trying to decide whether to keep the current stack or change course. Main question is how people here are handling workstation thermals.

Would appreciate practical advice rather than theory.

For context, I'm not chasing leaderboard numbers. I mostly want the system to feel consistent day to day.

Posting this in case it saves someone else a little time.

I'm helping a small team set up a local workflow. We don't need a giant stack, but we do care about MoE serving.

Would rather learn from your scars than collect my own all at once.

Did a bit of testing and thought I'd post the result here.

Short version: PyTorch behaved well for me, but only after I stopped ignoring quant drift.

Nothing scientific here, just practical notes from a real setup.

Maybe I'm missing something obvious.

I'm trying to decide whether to keep the current stack or change course. Main question is how people here are handling batch size.

If you've tested something similar, what did you keep and what did you stop bothering with?

Thought I'd post a small recipe because a lot of threads ask similar things.

Maybe useful as a starting point: vLLM on Threadripper box, keeping HBM constraints in mind from the start.

If someone wants the exact settings I can clean them up and post them.

This is mostly for local work rather than public serving.

Did a bit of testing and thought I'd post the result here.

Short version: TensorRT-LLM behaved well for me, but only after I stopped ignoring quant drift.

Happy to post more details if anyone wants them.

For context, I'm not chasing leaderboard numbers. I mostly want the system to feel consistent day to day.