DiffHLS: Differential Learning for High-Level Synthesis QoR Prediction with GNNs and LLM Code Embeddings
This work addresses the time-consuming synthesis process in HLS for hardware designers, presenting an incremental improvement by combining GNNs and LLM embeddings for more efficient optimization.
The paper tackles the expensive exploration of pragma-driven optimization choices in High-Level Synthesis (HLS) by proposing DiffHLS, a differential learning framework that predicts Quality-of-Result (QoR) from kernel-design pairs, achieving lower average MAPE than GNN baselines on PolyBench and showing scalability on ForgeHLS.
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.