DiffHLS: Differential Learning for High-Level Synthesis QoR Prediction with GNNs and LLM Code Embeddings
Zedong Peng, Zeju Li, Qiang Xu, Jieru Zhao
TLDR
DiffHLS uses differential learning with GNNs and LLM code embeddings to predict High-Level Synthesis Quality-of-Result faster.
Key contributions
- Introduces DiffHLS, a differential learning framework for HLS QoR prediction.
- Learns from kernel-design pairs, predicting baseline and design-induced delta for efficiency.
- Combines GNNs for IR graph encoding with LLM code embeddings for the delta pathway.
- Outperforms GNN baselines on PolyBench and demonstrates scalability on ForgeHLS.
Why it matters
High-Level Synthesis (HLS) optimization is currently slow and expensive. DiffHLS provides a faster, more accurate method for predicting Quality-of-Result, significantly accelerating design space exploration. This advancement makes HLS more practical and efficient for complex hardware development.
Original Abstract
High-Level Synthesis (HLS) compiles C/C++ into RTL, but exploring pragma-driven optimization choices remains expensive because each design point requires time-consuming synthesis. We propose \textbf{\DiffHLS}, a differential learning framework for HLS Quality-of-Result (QoR) prediction that learns from kernel--design pairs: a kernel baseline and a pragma-inserted design variant. \DiffHLS~encodes kernel and design intermediate-representation graphs with dedicated graph neural network (GNN) branches, and augments the delta pathway with code embeddings from a pretrained code large language model (LLM). Instead of regressing absolute targets directly, we jointly predict the kernel baseline and the design-induced delta, and compose them to obtain the design prediction. On PolyBench, \DiffHLS~attains lower average MAPE than GNN baselines under four GNN backbones, and LLM code embeddings consistently improve over a GNN-only ablation. We further validate scalability on the ForgeHLS dataset.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.