LIFT: LLM-Based Pragma Insertion for HLS via GNN Supervised Fine-Tuning
Computer Science Department, University of California – Los Angeles, USA
arXiv:2504.21187 [cs.LG], (29 Apr 2025)
@misc{prakriya2025liftllmbasedpragmainsertion,
title={LIFT: LLM-Based Pragma Insertion for HLS via GNN Supervised Fine-Tuning},
author={Neha Prakriya and Zijian Ding and Yizhou Sun and Jason Cong},
year={2025},
eprint={2504.21187},
archivePrefix={arXiv},
primaryClass={cs.LG},
url={https://arxiv.org/abs/2504.21187}
}
FPGAs are increasingly adopted in datacenter environments for their reconfigurability and energy efficiency. High-Level Synthesis (HLS) tools have eased FPGA programming by raising the abstraction level from RTL to untimed C/C++, yet attaining high performance still demands expert knowledge and iterative manual insertion of optimization pragmas to modify the microarchitecture. To address this challenge, we propose LIFT, a large language model (LLM)-based coding assistant for HLS that automatically generates performance-critical pragmas given a C/C++ design. We fine-tune the LLM by tightly integrating and supervising the training process with a graph neural network (GNN), combining the sequential modeling capabilities of LLMs with the structural and semantic understanding of GNNs necessary for reasoning over code and its control/data dependencies. On average, LIFT produces designs that improve performance by 3.52x and 2.16x than prior state-of the art AutoDSE and HARP respectively, and 66x than GPT-4o.
May 4, 2025 by hgpu
