repoGitLabTrust 82 · PrimaryPublished 10d agoLive · 9d ago
test5630352/llm-cache-optimize
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- PossiblePossibly related (embedding) · 62%I compiled LLM inference pricing across 7 providers — the caching numbers are surprising(spreadsheet included) [R] →
- PossiblePossibly related (embedding) · 53%DeepSeek-V4-Flash (MXFP4): compute buffer scales ~3x just from KV cache quant type (f16 vs q8_0) — anyone else seeing this? Llama.cpp →
- PossiblePossibly related (embedding) · 53%Llama-Server is Throwing Away Your Perfectly Good KV Caches, and How to Fix It →
- PossiblePossibly related (embedding) · 53%My DeepSeek V4 Pro at home got faster again →
- PossiblePossibly related (embedding) · 51%I mapped which local LLMs actually fit each RAM tier, 8 to 128GB (open dataset) →
- PossiblePossibly related (embedding) · 47%Recent llama.cpp updates for SYCL/Intel →
Covers
newsI compiled LLM inference pricing across 7 providers — the caching numbers are surprising(spreadsheet included) [R]newsDeepSeek-V4-Flash (MXFP4): compute buffer scales ~3x just from KV cache quant type (f16 vs q8_0) — anyone else seeing this? Llama.cppnewsLlama-Server is Throwing Away Your Perfectly Good KV Caches, and How to Fix ItnewsMy DeepSeek V4 Pro at home got faster againnewsI mapped which local LLMs actually fit each RAM tier, 8 to 128GB (open dataset)
Covers (incoming)
Related across the graph
newsI compiled LLM inference pricing across 7 providers — the caching numbers are surprising(spreadsheet included) [R]newsMy DeepSeek V4 Pro at home got faster againnewsRecent llama.cpp updates for SYCL/IntelnewsI mapped which local LLMs actually fit each RAM tier, 8 to 128GB (open dataset)newsDeepSeek-V4-Flash (MXFP4): compute buffer scales ~3x just from KV cache quant type (f16 vs q8_0) — anyone else seeing this? Llama.cppnewsLlama-Server is Throwing Away Your Perfectly Good KV Caches, and How to Fix It