driver/framework news roundup

[audiopoint18]

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 speculative decoding.

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

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

[amberrunner]

Appreciate the detail. I ended up in about the same place using vLLM. The biggest change for me was simplifying the stack instead of adding one more optimization layer.

[deepseeked]

Can you fill in a bit more context? Do you see the same thing with a smaller batch or shorter prompt?