Feed
RoofLang: Enabling AI-Driven Architecting of LLM Inference Systems
A simulator language lets an AI agent redesign LLM serving setups without running real code.
39.5×
Peak decode throughput gap
DeepSeek V4 Flash versus Kimi K3 at 1M context
How it works
Profiling hits limits
AI optimizers tune existing software by profiling it, so designs the stack can't run well never get found.
Simulate architectures instead
RoofLang describes workloads and hardware as graphs, allows only meaning-preserving rewrites, and scores them with a roofline simulator.
Insights and new designs
It shows compact KV caches drive DeepSeek V4's speed; a Claude Code agent found layouts up to 50.1% better.
From the paper