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

Short version: Ollama behaved well for me, but only after I stopped ignoring prefill/decode split.

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.

This keeps coming up in conversation so I figured I'd start a thread.

It feels like the conversation has shifted hard toward inference and serving details lately, and honestly that makes sense once you start dealing with small local stacks.

Wondering if this is just where the whole space is heading now.

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

I'm not trying to build a perfect architecture diagram. I just want a sane starting point around SGLang and mini PC.

Simple and boring is a valid answer. The logs were somehow calmer than I was.

I figured I'd ask here before changing everything.

My current setup is Ollama with 5090, and I keep running into continuous batching.

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

This keeps coming up in conversation so I figured I'd start a thread.

One thing I keep noticing is that people underestimate MoE serving until they actually try to run this stuff for more than one afternoon.

Maybe obvious to some of you, but it clicked for me this week.

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

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

Two paths both seem viable, but they break down differently once VRAM headroom becomes important.

Trying to avoid rebuilding the whole thing twice.

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

After a few runs on 4090, I ended up with decent performance, but the setup only really made sense once I adjusted around long context.

Curious if anyone else landed in the same place.

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

I've gone back and forth between PyTorch and LM Studio for Qwen. The interesting difference for me has been tail latency.

What would you pick if you had to live with it for a few months?

This keeps coming up in conversation so I figured I'd start a thread.

One thing I keep noticing is that people underestimate structured output until they actually try to run this stuff for more than one afternoon.

Curious whether that lines up with what other people are seeing.

This is mostly for local work rather than public serving.

Has anyone else run into this?

I'm mostly working with Mixtral in TensorRT-LLM on GB200. The part I'm trying to understand is batch size.

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.