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

Lorenzo Ferretti,Jihye Kwon,Giovanni Ansaloni,Giuseppe Di Guglielmo,Luca Carloni,Laura Pozzi

DOI: https://doi.org/10.48550/arXiv.2101.00587

2021-01-03

Abstract:High-Level Synthesis (HLS) frameworks allow to easily specify a large number of variants of the same hardware design by only acting on optimization directives. Nonetheless, the hardware synthesis of implementations for all possible combinations of directive values is impractical even for simple designs. Addressing this shortcoming, many HLS Design Space Exploration (DSE) strategies have been proposed to devise directive settings leading to high-quality implementations while limiting the number of synthesis runs. All these works require considerable efforts to validate the proposed strategies and/or to build the knowledge base employed to tune abstract models, as both tasks mandate the syntheses of large collections of implementations. Currently, such data gathering is performed ad-hoc, a) leading to a lack of standardization, hampering comparisons between DSE alternatives, and b) posing a very high burden to researchers willing to develop novel DSE strategies. Against this backdrop, we here introduce DB4HLS, a database of exhaustive HLS explorations comprising more than 100000 design points collected over 4 years of synthesis time. The open structure of DB4HLS allows the incremental integration of new DSEs, which can be easily defined with a dedicated domain-specific language. We think that of our database, available at <a class="link-external link-https" href="https://www.db4hls.inf.usi.ch/" rel="external noopener nofollow">this https URL</a>, will be a valuable tool for the research community investigating automated strategies for the optimization of HLS-based hardware designs.

Hardware Architecture

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to effectively reduce the required time and resources in High - Level Synthesis (HLS) Design Space Exploration (DSE), while ensuring that the performance - optimal design scheme can be found. Specifically, the paper points out:

1. **The HLS framework allows a large number of hardware design variants to be easily defined through optimized instructions**, but it is impractical to perform hardware synthesis for all possible combinations of instruction values in actual operation, even for simple designs. This leads to inefficiency in design space exploration and waste of resources.

2. **Although existing DSE strategies can find high - quality implementation schemes by limiting the number of syntheses**, a great deal of effort is required to synthesize a large number of implementations when verifying these strategies or building a knowledge base for adjusting abstract models. This data collection is usually ad - hoc, lacks standardization, hinders the comparison between different DSE methods, and places a heavy burden on researchers.

To address these problems, the paper introduces **DB4HLS**, an HLS design space exploration database containing more than 100,000 design points. This database aims to provide a standardized data set for easy comparison and evaluation of DSE strategies, while lowering the research threshold for developing new DSE strategies. By providing a rich design space and related synthesis results, DB4HLS helps promote research on automated HLS design optimization strategies.