do small edge boxes make sense with vLLM

[audionode8]

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

The more I work with local setups, the more it seems like VRAM headroom decides everything downstream.

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

[routecrate]

A couple things I'd check first. Before rebuilding anything, I'd test one clean baseline with Phi and compare from there.