PASS: exploiting post-activation sparsity in streaming architectures for CNN acceleration
File(s)
Accepted version
Author(s)
Montgomerie-Corcoran, Alexander
Yu, Zhewen
Cheng, Jianyi
Bouganis, Christos-Savvas
Type
Conference Paper
Abstract
With the ever-growing popularity of Artificial Intelligence, there is an increasing demand for more performant and efficient underlying hardware. Convolutional Neural Networks (CNN) are a workload of particular importance, which achieve high accuracy in computer vision applications. Inside CNNs, a significant number of the post-activation values are zero, resulting in many redundant computations. Recent works have explored this post-activation sparsity on instruction-based CNN accelerators but not on streaming CNN accelerators, despite the fact that streaming architectures are considered the leading design methodology in terms of performance. In this paper, we highlight the challenges associated with exploiting post-activation sparsity for performance gains in streaming CNN accelerators, and demonstrate our approach to address them. Using a set of modern CNN benchmarks, our streaming sparse accelerators achieve 1.41 x to 1.93 x efficiency (GOP/sDSP) compared to state-of-the-art instruction-based sparse accelerators.
Date Issued
2023-09-04
Date Acceptance
2023-09-04
Citation
2023 33rd International Conference on Field-Programmable Logic and Applications (FPL), 2023, pp.288-293
ISBN
979-8-3503-4151-5
Publisher
IEEE
Start Page
288
End Page
293
Journal / Book Title
2023 33rd International Conference on Field-Programmable Logic and Applications (FPL)
Copyright Statement
Copyright © 2023 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.
Identifier
http://dx.doi.org/10.1109/fpl60245.2023.00049
Source
2023 33rd International Conference on Field-Programmable Logic and Applications (FPL)
Publication Status
Published
Start Date
2023-09-04
Finish Date
2023-09-08
Coverage Spatial
Gothenburg, Sweden
Date Publish Online
2023-09-04