用大模型代理自动优化硬件参数，提速并提升性能。
LLM-DSE: Searching Accelerator Parameters with LLM Agents
- 构建四代理系统协同探索硬件参数空间
- 在HLSyn数据集上实现2.55倍性能提升
- 适合硬件设计与AI加速器研发人员
尽管高层次综合（HLS）工具通过提高抽象层次缓解了领域专用加速器（DSA）编程的挑战，但优化硬件指令参数仍是重大难题。现有启发式和基于学习的方法在适应性和样本效率方面表现不佳。我们提出LLM-DSE，一种专为优化HLS指令设计的多智能体框架。该框架结合大语言模型与设计空间探索（DSE），由路由器、专家、仲裁者和批评者四个智能体协同工作，与多种工具交互以加速优化过程。LLM-DSE利用关键领域知识识别高效参数组合，并通过在线互动中的自然语言学习保持灵活性。在HLSyn数据集上的评估显示，相比最先进方法，其性能提升达2.55倍，同时发现新设计并缩短运行时间。消融实验验证了各智能体协作的有效性与必要性。代码已开源：https://github.com/Nozidoali/LLM-DSE。
原文摘要 · Abstract (English)
Even though high-level synthesis (HLS) tools mitigate the challenges of programming domain-specific accelerators (DSAs) by raising the abstraction level, optimizing hardware directive parameters remains a significant hurdle. Existing heuristic and learning-based methods struggle with adaptability and sample efficiency. We present LLM-DSE, a multi-agent framework designed specifically for optimizing HLS directives. Combining LLM with design space exploration (DSE), our explorer coordinates four agents: Router, Specialists, Arbitrator, and Critic. These multi-agent components interact with various tools to accelerate the optimization process. LLM-DSE leverages essential domain knowledge to identify efficient parameter combinations while maintaining adaptability through verbal learning from online interactions. Evaluations on the HLSyn dataset demonstrate that LLM-DSE achieves substantial $2.55\times$ performance gains over state-of-the-art methods, uncovering novel designs while reducing runtime. Ablation studies validate the effectiveness and necessity of the proposed agent interactions. Our code is open-sourced here: https://github.com/Nozidoali/LLM-DSE.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。