Performance and Energy Aware Inhomogeneous 3D Networks-on-Chip Architecture Generation

Michael Opoku Agyeman, Ali Ahmadinia, Nader Bagherzadeh

Research output: Contribution to journalArticle

Abstract

Recently, Through-Silicon-Via (TSV) has been more popular to provide faster inter-layer communication in
three-dimensional Networks-on-Chip (3D NoCs). However, the area overhead of TSVs reduces wafer utilization and yield which
impact design of 3D architectures using a large number of TSVs such as homogeneous 3D NoCs topologies. Also, 3D routers require
more memory and thus they are more power hungry than conventional 2D routers. Alternatively, hybrid 3D NoCs combine both the area
and performance benefits of 2D and 3D router architectures by using a limited number of TSVs. Existing hybrid architectures suffer
from higher packet delays as they do not consider the dynamic communication patterns of different application and their NoC resource
usage. We propose a novel algorithm to systematically generate hybrid 3D NoC topologies for a given application such that the vertical
connections are minimized while the NoC performance is not sacrificed. The proposed algorithm analyses the target application and
generates hybrid architectures by efficiently redistributing the vertical links and buffer spaces based on their utilizations. Furthermore,
the algorithm has been evaluated with synthetic and various real-world traffic patterns. Experimental results show that the proposed
algorithm generates optimized architectures with lower energy consumption and a significant reduction in packet delay compared to the
existing solutions.
Original languageEnglish
JournalIEEE Transactions on Parallel and Distributed Systems
Publication statusAccepted/In press - 1 May 2016

Fingerprint

Routers
Topology
Communication
Energy utilization
Data storage equipment
Silicon
Network-on-chip

Keywords

  • Multi-Processor System
  • Network-on-Chip
  • 3D Integration
  • Performance Evaluation

Cite this

@article{25b71febf2f9492ca38fb749150afab6,
title = "Performance and Energy Aware Inhomogeneous 3D Networks-on-Chip Architecture Generation",
abstract = "Recently, Through-Silicon-Via (TSV) has been more popular to provide faster inter-layer communication inthree-dimensional Networks-on-Chip (3D NoCs). However, the area overhead of TSVs reduces wafer utilization and yield whichimpact design of 3D architectures using a large number of TSVs such as homogeneous 3D NoCs topologies. Also, 3D routers requiremore memory and thus they are more power hungry than conventional 2D routers. Alternatively, hybrid 3D NoCs combine both the areaand performance benefits of 2D and 3D router architectures by using a limited number of TSVs. Existing hybrid architectures sufferfrom higher packet delays as they do not consider the dynamic communication patterns of different application and their NoC resourceusage. We propose a novel algorithm to systematically generate hybrid 3D NoC topologies for a given application such that the verticalconnections are minimized while the NoC performance is not sacrificed. The proposed algorithm analyses the target application andgenerates hybrid architectures by efficiently redistributing the vertical links and buffer spaces based on their utilizations. Furthermore,the algorithm has been evaluated with synthetic and various real-world traffic patterns. Experimental results show that the proposedalgorithm generates optimized architectures with lower energy consumption and a significant reduction in packet delay compared to theexisting solutions.",
keywords = "Multi-Processor System, Network-on-Chip, 3D Integration, Performance Evaluation",
author = "{Opoku Agyeman}, Michael and Ali Ahmadinia and Nader Bagherzadeh",
year = "2016",
month = "5",
day = "1",
language = "English",
journal = "IEEE Transactions on Parallel and Distributed Systems",
issn = "1045-9219",
publisher = "Institute of Electrical and Electronics Engineers",

}

Performance and Energy Aware Inhomogeneous 3D Networks-on-Chip Architecture Generation. / Opoku Agyeman, Michael; Ahmadinia, Ali; Bagherzadeh, Nader.

In: IEEE Transactions on Parallel and Distributed Systems, 01.05.2016.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Performance and Energy Aware Inhomogeneous 3D Networks-on-Chip Architecture Generation

AU - Opoku Agyeman, Michael

AU - Ahmadinia, Ali

AU - Bagherzadeh, Nader

PY - 2016/5/1

Y1 - 2016/5/1

N2 - Recently, Through-Silicon-Via (TSV) has been more popular to provide faster inter-layer communication inthree-dimensional Networks-on-Chip (3D NoCs). However, the area overhead of TSVs reduces wafer utilization and yield whichimpact design of 3D architectures using a large number of TSVs such as homogeneous 3D NoCs topologies. Also, 3D routers requiremore memory and thus they are more power hungry than conventional 2D routers. Alternatively, hybrid 3D NoCs combine both the areaand performance benefits of 2D and 3D router architectures by using a limited number of TSVs. Existing hybrid architectures sufferfrom higher packet delays as they do not consider the dynamic communication patterns of different application and their NoC resourceusage. We propose a novel algorithm to systematically generate hybrid 3D NoC topologies for a given application such that the verticalconnections are minimized while the NoC performance is not sacrificed. The proposed algorithm analyses the target application andgenerates hybrid architectures by efficiently redistributing the vertical links and buffer spaces based on their utilizations. Furthermore,the algorithm has been evaluated with synthetic and various real-world traffic patterns. Experimental results show that the proposedalgorithm generates optimized architectures with lower energy consumption and a significant reduction in packet delay compared to theexisting solutions.

AB - Recently, Through-Silicon-Via (TSV) has been more popular to provide faster inter-layer communication inthree-dimensional Networks-on-Chip (3D NoCs). However, the area overhead of TSVs reduces wafer utilization and yield whichimpact design of 3D architectures using a large number of TSVs such as homogeneous 3D NoCs topologies. Also, 3D routers requiremore memory and thus they are more power hungry than conventional 2D routers. Alternatively, hybrid 3D NoCs combine both the areaand performance benefits of 2D and 3D router architectures by using a limited number of TSVs. Existing hybrid architectures sufferfrom higher packet delays as they do not consider the dynamic communication patterns of different application and their NoC resourceusage. We propose a novel algorithm to systematically generate hybrid 3D NoC topologies for a given application such that the verticalconnections are minimized while the NoC performance is not sacrificed. The proposed algorithm analyses the target application andgenerates hybrid architectures by efficiently redistributing the vertical links and buffer spaces based on their utilizations. Furthermore,the algorithm has been evaluated with synthetic and various real-world traffic patterns. Experimental results show that the proposedalgorithm generates optimized architectures with lower energy consumption and a significant reduction in packet delay compared to theexisting solutions.

KW - Multi-Processor System

KW - Network-on-Chip

KW - 3D Integration

KW - Performance Evaluation

M3 - Article

JO - IEEE Transactions on Parallel and Distributed Systems

JF - IEEE Transactions on Parallel and Distributed Systems

SN - 1045-9219

ER -