Hlsclaw speeds up hardware code optimization with smart ai agents
HLSClaw: Bringing Vibe Coding to High-Level Synthesis through Skill-Based Agentic Harness
Machine LearningSoftware Engineering
Summary
Making computer chips involves writing special code that needs careful tweaking to run fast and use less power. The authors found that typical AI tools struggle when trying to improve this special code. They created HLSClaw, a new AI system that breaks down the problem into small tasks and learns how to adjust code and settings together. This approach made the chip designs run much faster and better than existing methods.
What this means in practice
- •For hardware engineers: Optimize hardware description code more effectively by automatically balancing code changes and synthesis options to improve performance metrics.
- •For software developers for embedded systems: Accelerate embedded system development by integrating an agent-based AI that tunes hardware synthesis without manual trial and error.
Authors
Jinming Ge, Shang Liu, Jiawei Liang, Shangkun Li, Afzal Ahmad, Jiang Xu, Linfeng Du, Wei Zhang
Abstract
LLM agents are rapidly shifting software development toward intent-driven \emph{vibe coding}, yet their effectiveness drops sharply in high-level synthesis (HLS), where code optimization must jointly consider design structure, directive search, and downstream hardware constraints. Existing HLS optimization approaches, including both heuristic-based methods and recent LLM-assisted pipelines, typically operate at fixed abstraction levels or in weakly coupled stages, limiting their ability to leverage the strong interplay between source code restructuring and directive search outcomes. Our key insight is that HLS optimization can be reframed from a fixed, hardwired transformation pipeline into a skill-based agentic control problem: an LLM agent equipped with modular, executable optimization skills and a feedback-driven orchestrator can autonomously navigate the coupled space of code restructuring and directive search. This motivates the design of HLSClaw, the first HLS-specific, skill-based agentic harness for holistic HLS optimization. On real-world HLS benchmarks, HLSClaw achieves a 39.12x geometric-mean latency speedup, outperforming the compiler baseline ScaleHLS by 1.98x, and surpassing the commercial coding agents Claude Code and Codex by 1.86x and 1.70x, respectively, while maintaining strong timing robustness.