Prefill - GoKawiil

Viiwork tests KV-cache sharing across two RTX 5090 hosts over 1GbE vs 10GbE

A benchmark compared viiwork's kv_share feature on two RTX 5090 machines connected via 1 GbE, a direct 10 GbE cable running Tailscale, and the same 10 GbE cable sending kv_share traffic directly. The test ran a coding-agent workload building a todo app, measuring 12-session and 16-session runs over 90 minutes to see how network speed affects whether a session's cached prompt reaches a second host before it's needed.

viiwork.pulu.app · 2026-10-27

DeepSeek unveils V4.1 Flash with 4x KV cache compression and 420 tokens/sec speed

DeepSeek released V4.1 Flash, a model initially mistaken for a minor update but revealed via its technical report to be a significant architectural overhaul, effectively a V5-class release. It achieves near 420 tokens/second throughput while compressing KV cache by 4x through techniques including cross-layer compression, sparse attention indexing optimizations, and FP4 precision, alongside a YOCO-inspired prefill design that only activates 8B parameters during prefill versus 16B during decode across its 40 layers.

zartbot.github.io · 2026-09-17