DFlash Parallel Drafting Breaks Speculative Decoding Bottleneck

Zhijian Liu of incoai presented DFlash at Runtime, which drafts entire token blocks in parallel instead of one token at a time, accepting 21% more tokens with only 1.3% added latency.

2026-11-06 ~ 2026-11-06 · 2 related posts

• DFlash 2 parallel drafting lifts speculative decoding acceptance 21% for 1.3% latency cost — songhan_mit · 2026-11-06
• DFlash drafts whole token blocks in parallel to fix speculative decoding bottleneck — Lianhuiq · 2026-11-06