Gen-TAS: A Generative AI-Aided Hardware-Software Task Allocation Framework for FPGA-GPP Heterogeneous Systems
Gen-TAS, a knowledge-grounded LLM framework for user-specific FPGA-GPP task allocation, is proposed by combining task-graph analysis with RAG, which grounds LLM reasoning in historical implementation knowledge and generates multiple explainable strategies tailored to the specified objectives.
Mary Kong, Yu-Qin Zhao, Semih Vazgecen et al.
· 0 citations