Close

Session

Research Manuscript
:
Foundations and Frontiers: Bridging Traditional EDA and Generative AI
DescriptionAs the complexity of hardware design reaches unprecedented scales, the tools we rely on must evolve to keep pace. This session explores the intersection of foundational EDA techniques and emerging machine learning paradigms. The session focuses on integrating Large Language Models and Machine Learning techniques to automate and optimize High-Level Synthesis, Design Space Exploration, and Verilog code generation for FPGAs.
Event Type
Research Manuscript
TimeMonday, July 2710:30am - 12:30pm PDT
LocationMtg Room 201B
Topics
EDA
Tracks
EDA5. RTL/Logic Level and High-level Synthesis
Presentations
10:30am - 10:43am PDTCHiRM-DSE: CodeLLM-Based Hierarchical and Rule-Mining Guided DSE for FPGA Accelerators
10:43am - 10:56am PDTIntegrating LLMs into a Two-Tier Traditional Methodology for Automated Verilog Code Generation
10:56am - 11:10am PDTLLM-Based Automatic Architecture Optimization for AI Model HLS Implementation on FPGA
11:10am - 11:23am PDTInsights from Verification: Training a Verilog Generation LLM Using Reinforcement Learning with Testbench Feedback
11:23am - 11:36am PDTUnlocking Automated Datapath Gating via Machine Learning Power Prediction
11:36am - 11:50am PDTSoberdse: Sample-Efficient Design Space Exploration via Learning-Based Algorithm Selection
11:50am - 12:03pm PDTEfficient HLS Accelerator Floorplan on Multi-Die FPGA Aided by Graph Neural Networks
12:03pm - 12:16pm PDTB-Flex: Exploration of Broader Flip-Flop Design Space Based on FSM Exhaustive Search
12:16pm - 12:30pm PDTHierarchical Boundary Recovery: Overcoming Synthesis Obscurity via SAT Sweeping