Unverified50% confidenceSolutionExact time
Setting kv=1 in GQA gives each layer only one key-value vector, compressing the KV cache to be almost independent of sequence length, thereby avoiding crowding out L3 space.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedKV Cache memory usage is proportional to the number of layers × number of attention heads × context length × precision; GQA reduces cache size by having multiple query heads share the same set of Key-Value heads, and this design has been adopted by the Llama 2/3 and Qwen series.69% similarUnverified使用非GQA架构（KV头数等于查询头数）在相同条件下的KV缓存是GQA的4倍68% similarUnverifiedGQA allows multiple query heads to share the same set of KV heads, compressing the KV cache size to 1/4 to 1/8 of the original.67% similarUnverifiedRL3通过状态离散化控制Q表规模，将连续状态空间分桶映射为有限离散集合以在浏览器中运行66% similarUnverifiedThe core bottleneck in large model inference on GPUs is memory bandwidth; autoregressive generation of each token requires loading the full KV Cache from HBM, with compute utilization typically below 5%.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1006042API
curl https://kongchang.com/api/v1/knowledge/claims/1006042MCP
get_claim(id=1006042)