MUGNoC: a software-configured multicast-unicast-gather NoC for accelerating CNN dataflows

Current communication infrastructures for convolutional neural networks (CNNs) only focus on specific transmission patterns, not applicable to benefit the whole system if the dataflow changes or different dataflows run in one system. To reduce data movement, various CNN dataflows are presented. For...

Full description

Saved in:
Bibliographic Details
Main Authors: Chen, Hui, Liu, Di, Li, Shiqing, Huai, Shuo, Luo, Xiangzhong, Liu, Weichen
Other Authors: School of Computer Science and Engineering
Format: Conference or Workshop Item
Language:English
Published: 2023
Subjects:
Online Access:https://hdl.handle.net/10356/165562
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-165562
record_format dspace
spelling sg-ntu-dr.10356-1655622023-05-26T15:35:50Z MUGNoC: a software-configured multicast-unicast-gather NoC for accelerating CNN dataflows Chen, Hui Liu, Di Li, Shiqing Huai, Shuo Luo, Xiangzhong Liu, Weichen School of Computer Science and Engineering 28th Asia and South Pacific Design Automation Conference (ASPDAC 2023) Engineering::Computer science and engineering::Hardware CNN Dataflow Parallel Multipath Transmission Network-on-Chips Current communication infrastructures for convolutional neural networks (CNNs) only focus on specific transmission patterns, not applicable to benefit the whole system if the dataflow changes or different dataflows run in one system. To reduce data movement, various CNN dataflows are presented. For these dataflows, parameters and results are delivered using different traffic patterns, i.e., multicast, unicast, and gather, preventing dataflow-specific communication backbones from benefiting the entire system if the dataflow changes or different dataflows run in the same system. Thus, in this paper, we propose MUG-NoC to support typical traffic patterns and accelerate them, therefore boosting multiple dataflows. Specifically, (i) we for the first time support multicast in 2D-mesh software configurable NoC by revising router configuration and proposing the efficient multicast routing; (ii) we decrease unicast latency by transmitting data through the different routes in parallel; (iii) we reduce output gather overheads by pipelining basic dataflow units. Experiments show that at least our proposed design can reduce 39.2% total data transmission time compared with the state-of-the-art CNN communication backbone. Ministry of Education (MOE) Nanyang Technological University Submitted/Accepted version This work is partially supported by the Ministry of Education, Singapore, under its Academic Research Fund Tier 2 (MOE2019-T2- 1-071) and Tier 1 (MOE2019-T1-001-072), and partially supported by Nanyang Technological University, Singapore, under its NAP (M4082282). 2023-03-31T05:41:41Z 2023-03-31T05:41:41Z 2023 Conference Paper Chen, H., Liu, D., Li, S., Huai, S., Luo, X. & Liu, W. (2023). MUGNoC: a software-configured multicast-unicast-gather NoC for accelerating CNN dataflows. 28th Asia and South Pacific Design Automation Conference (ASPDAC 2023), 308-313. https://dx.doi.org/10.1145/3566097.3567846 https://hdl.handle.net/10356/165562 10.1145/3566097.3567846 308 313 en MOE2019-T2-1-071 MOE2019-T1-001-072 NAP (M4082282) © 2023 Association for Computing Machinery. All rights reserved. This paper was published in the Proceedings of the 28th Asia and South Pacific Design Automation Conference (ASPDAC 2023) and is made available with permission of Association for Computing Machinery. application/pdf
institution Nanyang Technological University
building NTU Library
continent Asia
country Singapore
Singapore
content_provider NTU Library
collection DR-NTU
language English
topic Engineering::Computer science and engineering::Hardware
CNN Dataflow
Parallel Multipath Transmission
Network-on-Chips
spellingShingle Engineering::Computer science and engineering::Hardware
CNN Dataflow
Parallel Multipath Transmission
Network-on-Chips
Chen, Hui
Liu, Di
Li, Shiqing
Huai, Shuo
Luo, Xiangzhong
Liu, Weichen
MUGNoC: a software-configured multicast-unicast-gather NoC for accelerating CNN dataflows
description Current communication infrastructures for convolutional neural networks (CNNs) only focus on specific transmission patterns, not applicable to benefit the whole system if the dataflow changes or different dataflows run in one system. To reduce data movement, various CNN dataflows are presented. For these dataflows, parameters and results are delivered using different traffic patterns, i.e., multicast, unicast, and gather, preventing dataflow-specific communication backbones from benefiting the entire system if the dataflow changes or different dataflows run in the same system. Thus, in this paper, we propose MUG-NoC to support typical traffic patterns and accelerate them, therefore boosting multiple dataflows. Specifically, (i) we for the first time support multicast in 2D-mesh software configurable NoC by revising router configuration and proposing the efficient multicast routing; (ii) we decrease unicast latency by transmitting data through the different routes in parallel; (iii) we reduce output gather overheads by pipelining basic dataflow units. Experiments show that at least our proposed design can reduce 39.2% total data transmission time compared with the state-of-the-art CNN communication backbone.
author2 School of Computer Science and Engineering
author_facet School of Computer Science and Engineering
Chen, Hui
Liu, Di
Li, Shiqing
Huai, Shuo
Luo, Xiangzhong
Liu, Weichen
format Conference or Workshop Item
author Chen, Hui
Liu, Di
Li, Shiqing
Huai, Shuo
Luo, Xiangzhong
Liu, Weichen
author_sort Chen, Hui
title MUGNoC: a software-configured multicast-unicast-gather NoC for accelerating CNN dataflows
title_short MUGNoC: a software-configured multicast-unicast-gather NoC for accelerating CNN dataflows
title_full MUGNoC: a software-configured multicast-unicast-gather NoC for accelerating CNN dataflows
title_fullStr MUGNoC: a software-configured multicast-unicast-gather NoC for accelerating CNN dataflows
title_full_unstemmed MUGNoC: a software-configured multicast-unicast-gather NoC for accelerating CNN dataflows
title_sort mugnoc: a software-configured multicast-unicast-gather noc for accelerating cnn dataflows
publishDate 2023
url https://hdl.handle.net/10356/165562
_version_ 1772828019968180224