do you even need a vector DB here: setup advice

[deltagrid]

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

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

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

This is mostly for local work rather than public serving.

[sparrowghost]

Can you fill in a bit more context? Are you testing this on B200 or something else?

[modgarden]

Tried to reproduce it just now. My result was mixed: first run looked bad, second run looked normal. The main thing that moved was latency, not quality.

[oxiderunner]

that fixed it for me too.

[routearch]

that fixed it for me too.

worth checking the driver first.

[cacheline]

That matches what I was seeing. I ended up in about the same place using llama.cpp. The biggest change for me was simplifying the stack instead of adding one more optimization layer.

[tensorfield88]

Good point. I ended up in about the same place using SGLang.

[macrotinker]

Not sure I agree there. I've seen people blame batch size when the real issue was much more mundane. Could be a workload difference, but I wouldn't generalize too fast from one setup.

[embersmith]

A couple things I'd check first. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at consumer GPUs. Once you know the plain path works, then add the faster path back in piece by piece.

[promptfield]

I hit almost the same issue last night.