Back
Ssignal
16
·9 days ago·1 signals
Archived topic · source no longer tracked

Auto-fit vs tuned MoE offload: 564 → 1330 pp tok/s, unchanged decode (Qwen3.6-35B-A3B Q6 / RTX 3090)

NVIDIAOn-device

Heat trend

Collecting trend data

The percentage is based on available heat signal, not comment count or independent people.