Four days ago I released a hyper-optimized Qwen3.8-27B inference engine for an RTX 3090 (82 tps single request, 672 peak). Since then it went to ~114, then ~138 tps single-user with DFlash2 drafting and lookup-augmented drafting. Today it's ~133 tps on real chat prompts, 382 tps when the model repro
