driver/framework news roundup on 4090

[vramghost]

A couple questions before I guess too hard. Are you testing this on Threadripper box or something else?

There are a few different conversations hiding inside this thread, so being explicit about the goal helps.

[bytelab]

same here on my side.

What worked for me was pretty simple. If you're on vLLM, I'd confirm the version first and then reduce variables until the behavior is obvious. Not glamorous advice, but repeatable baselines save a lot of time.

[tensorport]

same here on my side.

[amberrunner]

Not sure I agree there. 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.

If you post your exact settings, people here can probably narrow it down pretty fast.