Abstract:Despite the great success of high-level synthesis (HLS) tools, we observe several unresolved challenges: 1) the high-level abstraction of HLS programming styles sometimes conceals optimization opportunities; 2) the actual quality of resulting RTL designs is hard to predict; and 3) existing HLS tools do not provide flexible tradeoff (Pareto) solutions among different objectives and constraints. To this end, we propose an end-to-end framework, namely, IronMan-Pro. The primary goal is to enable a flexible and automated design space exploration (DSE), to provide either optimized solutions under user-specified constraints or Pareto tradeoffs among different objectives (such as resource types, area, and latency). IronMan-Pro consists of three components: 1) GPP, a highly accurate graph-neural-network-based performance and resource predictor; 2) RLMD, a reinforcement-learning-based multiobjective design exploration engine for optimal resource allocation strategies, aiming to provide Pareto solutions among different objectives; and 3) CT, a code transformer to assist RLMD and GPP, which extracts the data flow graphs from original HLS C/C++ and automatically generates synthesizable code with optimized HLS directives. Experimental results show that, 1) GPP achieves high prediction accuracy, reducing the prediction errors of HLS tools by in resource utilization and in critical path (CP) timing; 2) compared with meta-heuristic-based techniques, IronMan-Pro generates superior solutions improving resource utilization by and CP timing by 7.6%–16.5%; and - ) under user-specified constraints, IronMan-Pro can find satisfying solutions over 96% of the cases, more than twice as many as that of meta-heuristic-based techniques and with a speedup of up to . This work demonstrates the great potential of applying machine learning algorithms in the electronic design automation domain, especially for the hard-to-solve problems, such as timing estimation and optimization. IronMan-Pro is available at https://github.com/lydiawunan/IronMan.

Multi-objective Design Space Exploration for High-Level Synthesis Via Bayesian Optimization

HGBO-DSE: Hierarchical GNN and Bayesian Optimization Based HLS Design Space Exploration

Learning from the Past: Efficient High-level Synthesis Design Space Exploration for FPGAs

Deep Inverse Design for High-Level Synthesis

Practical Design Space Exploration

Chimera: A Hybrid Machine Learning Driven Multi-Objective Design Space Exploration Tool for FPGA High-Level Synthesis

IronMan-Pro: Multiobjective Design Space Exploration in HLS via Reinforcement Learning and Graph Neural Network-Based Modeling

AutoHLS: Learning to Accelerate Design Space Exploration for HLS Designs

Hierarchical Optimization Of Multi-Objective Embedded System Design Using Pareto-Frontier Algebra

Design space exploration of multiple loops on FPGAs using high level synthesis

System-Level Design Space Exploration for High-Level Synthesis under End-to-End Latency Constraints

Automated Design Space Exploration of Coarse-Grained Reconfigurable Architecture Via Bayesian Optimization

FlexWalker: an Efficient Multi-Objective Design Space Exploration Framework for HLS Design

Design Space Exploration in Multi-Objective Hierarchical SOC Design

A Particle Swarm Optimization-Based Method For Multiobjective Design Optimizations

Design Space Exploration of FPGA-based Accelerators with Multi-Level Parallelism

Multi-objective Optimization in CPU Design Space Exploration: Attention is All You Need

Ensemble of surrogate based global optimization methods using hierarchical design space reduction

DB4HLS: A Database of High-Level Synthesis Design Space Explorations

Graph Neural Networks for High-Level Synthesis Design Space Exploration

Microarchitecture Design Space Exploration Via Pareto-Driven Active Learning.