Posts: 52
Threads: 5
Joined: Aug 2026
Reputation:
0
Quote:[quote][quote]A couple things I'd check first. Before rebuilding anything, I'd test one clean baseline with Phi and compare from there. Not glamorous
Tried to reproduce it just now. I couldn't reproduce the slowdown until I pushed context much higher. The surprise was that memory pressure showed up before throughput collapsed.
trying not to rebuild again
Posts: 56
Threads: 7
Joined: Sep 2026
Reputation:
0
That matches what I was seeing. That matches what I saw on A6000. The biggest change for me was simplifying the stack instead of adding one more optimization layer.
Posts: 41
Threads: 7
Joined: Sep 2026
Reputation:
0
Good point. The pattern around consumer GPUs 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.
small models can still be good