Hybrid Graph Representation and Learning Framework for High-Level Synthesis Design Space Exploration
Optimizing hardware accelerators in high-level synthesis (HLS) relies on design space exploration (DSE), which involves experimenting with different pragma options and trading off hardware cost and performance metrics (HCPMs) to identify Pareto-optimal solutions. The exponential growth of the design...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10772109/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850123032966725632 |
|---|---|
| author | Pouya Taghipour Eric Granger Yves Blaquiere |
| author_facet | Pouya Taghipour Eric Granger Yves Blaquiere |
| author_sort | Pouya Taghipour |
| collection | DOAJ |
| description | Optimizing hardware accelerators in high-level synthesis (HLS) relies on design space exploration (DSE), which involves experimenting with different pragma options and trading off hardware cost and performance metrics (HCPMs) to identify Pareto-optimal solutions. The exponential growth of the design space, poor quality-of-results (QoR) estimation by HLS tools, and lengthy post-implementation runtime have made the HLS DSE process highly challenging and time-consuming. Automating this process could reduce time-to-market and associated development costs. Learning-based methods, particularly graph neural networks (GNNs), have shown considerable potential in addressing HLS QoR/DSE problems by modeling the mapping function from control data flow graphs (CDFGs) of HLS designs to their logic, enabling early estimation of QoR during the compilation phase of the hardware design flow. However, there is still a gap in terms of their prediction accuracy. Indeed, modeling HLS-related problems using GNNs that efficiently capture the complex patterns arising from applied pragmas and low-level characteristics of HLS specifications is challenging. This paper introduces a novel hybrid graph representation and learning framework under a multi-task setting, featuring two distinct types of CDFGs derived from two different sources. Furthermore, various models are proposed to fuse features and knowledge in joint, sequential, and parallel learning architectures, aiming to improve the overall accuracy and generalization in predicting QoR and approximating the Pareto frontier (PF). Experimental results show that our framework can attain a higher level of performance than the state-of-the-art baseline models over unseen designs, with an average relative improvement of 47.4 % and 16.0 % for resource utilization and performance metrics, respectively. Additionally, considering various HLS designs with different design space sizes, a 26.8 % enhancement in DSE PF approximation is observed. |
| format | Article |
| id | doaj-art-75c0eadd72d94c3c93ee39758723e435 |
| institution | OA Journals |
| issn | 2169-3536 |
| language | English |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-75c0eadd72d94c3c93ee39758723e4352025-08-20T02:34:42ZengIEEEIEEE Access2169-35362024-01-011218957418958910.1109/ACCESS.2024.350960610772109Hybrid Graph Representation and Learning Framework for High-Level Synthesis Design Space ExplorationPouya Taghipour0https://orcid.org/0000-0001-5166-2790Eric Granger1https://orcid.org/0000-0001-6116-7945Yves Blaquiere2https://orcid.org/0000-0001-6204-7427Department of Electrical Engineering, LaCIME, École de Technologie Supérieure (ÉTS), Montreal, QC, CanadaDepartment of Systems Engineering, LIVIA, ILLS, École de Technologie Supérieure (ÉTS), Montreal, QC, CanadaDepartment of Electrical Engineering, LaCIME, École de Technologie Supérieure (ÉTS), Montreal, QC, CanadaOptimizing hardware accelerators in high-level synthesis (HLS) relies on design space exploration (DSE), which involves experimenting with different pragma options and trading off hardware cost and performance metrics (HCPMs) to identify Pareto-optimal solutions. The exponential growth of the design space, poor quality-of-results (QoR) estimation by HLS tools, and lengthy post-implementation runtime have made the HLS DSE process highly challenging and time-consuming. Automating this process could reduce time-to-market and associated development costs. Learning-based methods, particularly graph neural networks (GNNs), have shown considerable potential in addressing HLS QoR/DSE problems by modeling the mapping function from control data flow graphs (CDFGs) of HLS designs to their logic, enabling early estimation of QoR during the compilation phase of the hardware design flow. However, there is still a gap in terms of their prediction accuracy. Indeed, modeling HLS-related problems using GNNs that efficiently capture the complex patterns arising from applied pragmas and low-level characteristics of HLS specifications is challenging. This paper introduces a novel hybrid graph representation and learning framework under a multi-task setting, featuring two distinct types of CDFGs derived from two different sources. Furthermore, various models are proposed to fuse features and knowledge in joint, sequential, and parallel learning architectures, aiming to improve the overall accuracy and generalization in predicting QoR and approximating the Pareto frontier (PF). Experimental results show that our framework can attain a higher level of performance than the state-of-the-art baseline models over unseen designs, with an average relative improvement of 47.4 % and 16.0 % for resource utilization and performance metrics, respectively. Additionally, considering various HLS designs with different design space sizes, a 26.8 % enhancement in DSE PF approximation is observed.https://ieeexplore.ieee.org/document/10772109/Electronic design automation (EDA)high-level synthesis (HLS)design space exploration (DSE)machine learning (ML)graph neural networks (GNN)field-programmable gate array (FPGA) |
| spellingShingle | Pouya Taghipour Eric Granger Yves Blaquiere Hybrid Graph Representation and Learning Framework for High-Level Synthesis Design Space Exploration IEEE Access Electronic design automation (EDA) high-level synthesis (HLS) design space exploration (DSE) machine learning (ML) graph neural networks (GNN) field-programmable gate array (FPGA) |
| title | Hybrid Graph Representation and Learning Framework for High-Level Synthesis Design Space Exploration |
| title_full | Hybrid Graph Representation and Learning Framework for High-Level Synthesis Design Space Exploration |
| title_fullStr | Hybrid Graph Representation and Learning Framework for High-Level Synthesis Design Space Exploration |
| title_full_unstemmed | Hybrid Graph Representation and Learning Framework for High-Level Synthesis Design Space Exploration |
| title_short | Hybrid Graph Representation and Learning Framework for High-Level Synthesis Design Space Exploration |
| title_sort | hybrid graph representation and learning framework for high level synthesis design space exploration |
| topic | Electronic design automation (EDA) high-level synthesis (HLS) design space exploration (DSE) machine learning (ML) graph neural networks (GNN) field-programmable gate array (FPGA) |
| url | https://ieeexplore.ieee.org/document/10772109/ |
| work_keys_str_mv | AT pouyataghipour hybridgraphrepresentationandlearningframeworkforhighlevelsynthesisdesignspaceexploration AT ericgranger hybridgraphrepresentationandlearningframeworkforhighlevelsynthesisdesignspaceexploration AT yvesblaquiere hybridgraphrepresentationandlearningframeworkforhighlevelsynthesisdesignspaceexploration |