DaqulaLin/MemStitch

Zero-Copy Context Bridging Gateway for Multi-Agent GPU Inference. Bypasses the expensive prefill phase by dynamically stitching KV Caches at the memory level using PagedAttention. Cuts TTFT latency by up to 25x and saves 40%+ VRAM for collaborative LLM workflows.

12 best points · 1 comments · 1 Hacker News post · 22 GitHub stars · View on GitHub

Hacker News discussions

• Show HN: MemStitch – Zero-copy context bridging for vLLM (25x TTFT speedup) (12 points, 1 comments, by daqulalin, 2026-07-14)