How Much KV Cache Budget Do We Need for LLM Serving?
Estimate the KV Cache capacity for LLM inference workloads by analyzing hit rate and prefill speedup under different cache budgets, with the help of KV Cache Hit Rate Simulator.
Jun 26, 2026
OpenClaw + Mooncake: A Stability Upgrade for Real-World Multi-Session Inference
By integrating Mooncake into OpenClaw's real inference path, we not only improved fast-path latency, but also sharply reduced TTFT tail latency in multi-session, long-context workloads, turning a system that was usually fast but occasionally slow into one that feels consistently smooth.
Mar 19, 2026
Mooncake Joins the PyTorch Ecosystem
Mooncake is now part of the PyTorch Ecosystem, complementing PyTorch-native LLM serving with high-performance disaggregated data transfer and storage.
Feb 12, 2026