Get Started
Home
Topics
Search
Library
Research questionHow can NL-to-HLS generation optimize hardware latency and resource use without costly synthesis-based training?Natural-language-to-HLS systems can produce functionally valid C/C++ while making pragma or structural choices that increase hardware latency or resource use. Repeated synthesis feedback is expensive during training, making QoR optimization difficult.
AI
Code Generation & Program Synthesis
Evaluation & Benchmarks
Reinforcement Learning
Research Paper
Technology
Latest papersRecent research connected to this question, newest first.HLS-Seek: QoR-Aware Code Generation for High-Level Synthesis via Proxy Comparative Reward Reinforcement LearningApplies to algorithmic C/C++ compiled into hardware through HLS, where QoR is measured by latency and resource utilization. The source evaluates a 7B-parameter system on HLS-Eval and 30 kernels, reporting syntax and functional correctness, latency, Pareto dominance, proxy comparison accuracy, and training speed; it specifically uses Vitis HLS for synthesis feedback.research paper · Sep 4, 2026
Related questions
How can LLMs synthesize executable machines and improve their designs through physical simulation feedback?How can LLM-generated GPU kernels remain performant across hardware, inputs, and programming models?How can LLMs reliably construct analytical hardware-performance models from workload, architecture, and mapping specifications?How can physical-aware logic synthesis improve PPA and closure time despite inaccurate wirelength estimates and large netlists?