Unverified50% confidenceFactExact time
In LLM inference, the decode phase is mainly limited by memory bandwidth, so a drop in SM frequency does not cause a proportional drop in speed; the prefill phase is compute-limited and is more severely affected by throttling.
1
Sources
50%
Confidence
Long-term
Relevance
10/17/2026
First Seen
Sources
Related Entities
Related Claims
Unverified当输出空间被DSL压缩后，可在一定程度上降低LLM幻觉的发生频率72% similarVerifiedThe prefill stage of Transformer inference is primarily bottlenecked by compute (FLOPS), while the decode stage is primarily bottlenecked by memory bandwidth.72% similarUnverifiedThe prefill stage is highly parallel with large throughput but latency that grows with context length, while the decode stage generates tokens auto-regressively and its speed is limited by memory bandwidth.71% similarUnverified只有在LLM推理比embedding加向量检索慢数十倍以上时，语义缓存才能同时实现降本与提速70% similarUnverified将高频连续控制信号映射到低带宽语言空间会引入LLM推理数十至数百毫秒延迟，对实时任务是致命缺陷70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1050214API
curl https://kongchang.com/api/v1/knowledge/claims/1050214MCP
get_claim(id=1050214)