what are you all running this week on GB200

[llm_wren]

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

Two paths both seem viable, but they break down differently once KV cache pressure becomes important.

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

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

[rocmplane]

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.

[vectorshed43]

that fixed it for me too.

[thetagrid15]

I think the bottleneck is somewhere else. I wouldn't optimize around that first. On my side the bigger swing came from the runtime and not the model. Worth testing the simpler explanation before adopting the more interesting one.

[ggufwire22]

Thanks, that helps. The pattern around quant drift has been pretty similar for me. Worth rechecking your defaults though, because one bad assumption can make the rest of the thread look confusing.

[marbletrack]

I'd approach it in a boring order. Before rebuilding anything, I'd test one clean baseline with Mistral and compare from there. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.

[rocmplane]

My first guess would be this: On dual 3090, 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.

[cacheplane98]

following this because I'm in the same boat.

[quiettoken]

My first guess would be this: Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at prefix caching. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.

[quiettoken]

That matches what I was seeing. The pattern around HBM constraints has been pretty similar for me. I got better results once I stopped chasing one metric and just looked at whether the workflow felt usable.