KV Prefix Cache 재사용 극대화로 TTFT 20~33% 단축한 CacheWeaver
CacheWeaver Reorders RAG Evidence for Prefix-Cache Reuse: Prefix-Cache-Aware Evidence Reordering
LPDDR5x 기반 480GB 메모리로 Prefill 가속화하는 Intel Crescent Island
Intel's mysterious new datacenter GPU is what Nvidia's Rubin CPX nearly was
Gemma 4 E4B: 128K Context Recall 완벽 구현 및 Prefill 지연 분석
I stress-tested Gemma 4 E4B's 128K context on a laptop GPU — recall is great, prefill is not
RAG의 Retrieval 과정을 제거한 Persistent KV Cache 기반 컨텍스트 최적화
We Replaced Our RAG Pipeline With Persistent KV Cache. Here's What We Found.
Prefill/Decode 분리 및 Optical Core 도입을 통한 AI Inference 최적화
Inference is giving AI chip startups a second chance to make their mark