Model reportedly uses QAT for KV cache, explaining strong fp4 performance while dropping MTP
stochasticchasm · x · 2026-09-11
An observer notes the new model applies quantization-aware training (QAT) to the KV cache as well, which explains its better-than-peers performance at fp4 KV cache. The author also noticed MTP (multi-token prediction) was removed, speculating the performance uplift wasn't ultimately worth it.
Related event: Leaked DeepSeek V4.1 Flash Tech Report Reveals KV Cache-Obsessed Flagship(44 posts)→
More from Infra
- DeepSpeed shows AutoEP/AutoTP/AutoSP parallelism gains, hosts office hours — PyTorch · 2026-10-27
- Running a local 9B as a coding agent: brief the big model to plan, let the small model execute — Electrical_Ninja3805 · 2026-10-27
- Cloudflare Durable Objects validated as top architecture for AI agent WebSocket servers — tristanbob · 2026-10-27
- Gary Marcus: With Q3 data added, Anthropic's compute-efficiency growth shows deceleration — GaryMarcus · 2026-10-27
- SpaceX locks in $40B GPU financing deal with Apollo, the largest ever committed — TechCrunch AI · 2026-10-27
- Inference is just Kubernetes with a multi-trillion-dollar market, and the communities overlap — matt_slotnick · 2026-10-27