Power Optimal Network-On-Chip Interconnect Design

A large part of today's multi-core chips is interconnect. Increasing communication complexity has made new strategies for interconnects essential such as Network on Chip. Power dissipation in interconnects has become a substantial part of the total power dissipation. Hence, techniques to reduce...

Full description

Bibliographic Details
Main Author: Vikas, G
Other Authors: Varghese, Kuruvilla
Language:en_US
Published: 2011
Subjects:
Online Access:http://etd.iisc.ernet.in/handle/2005/1408
http://etd.ncsi.iisc.ernet.in/abstracts/1817/G23700-Abs.pdf
id ndltd-IISc-oai-etd.ncsi.iisc.ernet.in-2005-1408
record_format oai_dc
spelling ndltd-IISc-oai-etd.ncsi.iisc.ernet.in-2005-14082018-01-10T03:35:55ZPower Optimal Network-On-Chip Interconnect DesignVikas, GNetwork On Chip - Design and ConstructionElectric Power NetworksApplication Specific System On ChipRouters (Computer Networks)Chip Multi Core ProcessorNetwork-on-Chip Interconnect DesignComputer EngineeringA large part of today's multi-core chips is interconnect. Increasing communication complexity has made new strategies for interconnects essential such as Network on Chip. Power dissipation in interconnects has become a substantial part of the total power dissipation. Hence, techniques to reduce interconnect power have become a necessity. In this thesis, we present a design methodology that gives values of bus width for interconnect links, frequency of operation for routers, in Network on Chip scenario that satisfy required throughput and dissipate minimal switching power. We develop closed form analytical expressions for the power dissipation, with bus width and frequency as variables and then use Lagrange multiplier method to arrive at the optimal values. To validate our methodology, we implement the router design in 90 nm technology and measure power for various bus widths and frequency combinations. We find that the experimental results are in good agreement with the predicted theoretical results. Further, we present the scenario of an Application Specific System on Chip (ASSoC), where the throughput requirements are different on different links. We show that our analytical model holds in this case also. Then, we present modified version of the solution considered for Chip Multi Processor (CMP) case that can solve the ASSoC scenario also.Varghese, KuruvillaKuri, Joy2011-09-08T05:40:56Z2011-09-08T05:40:56Z2011-09-082010-02Thesishttp://etd.iisc.ernet.in/handle/2005/1408http://etd.ncsi.iisc.ernet.in/abstracts/1817/G23700-Abs.pdfen_USG23700
collection NDLTD
language en_US
sources NDLTD
topic Network On Chip - Design and Construction
Electric Power Networks
Application Specific System On Chip
Routers (Computer Networks)
Chip Multi Core Processor
Network-on-Chip Interconnect Design
Computer Engineering
spellingShingle Network On Chip - Design and Construction
Electric Power Networks
Application Specific System On Chip
Routers (Computer Networks)
Chip Multi Core Processor
Network-on-Chip Interconnect Design
Computer Engineering
Vikas, G
Power Optimal Network-On-Chip Interconnect Design
description A large part of today's multi-core chips is interconnect. Increasing communication complexity has made new strategies for interconnects essential such as Network on Chip. Power dissipation in interconnects has become a substantial part of the total power dissipation. Hence, techniques to reduce interconnect power have become a necessity. In this thesis, we present a design methodology that gives values of bus width for interconnect links, frequency of operation for routers, in Network on Chip scenario that satisfy required throughput and dissipate minimal switching power. We develop closed form analytical expressions for the power dissipation, with bus width and frequency as variables and then use Lagrange multiplier method to arrive at the optimal values. To validate our methodology, we implement the router design in 90 nm technology and measure power for various bus widths and frequency combinations. We find that the experimental results are in good agreement with the predicted theoretical results. Further, we present the scenario of an Application Specific System on Chip (ASSoC), where the throughput requirements are different on different links. We show that our analytical model holds in this case also. Then, we present modified version of the solution considered for Chip Multi Processor (CMP) case that can solve the ASSoC scenario also.
author2 Varghese, Kuruvilla
author_facet Varghese, Kuruvilla
Vikas, G
author Vikas, G
author_sort Vikas, G
title Power Optimal Network-On-Chip Interconnect Design
title_short Power Optimal Network-On-Chip Interconnect Design
title_full Power Optimal Network-On-Chip Interconnect Design
title_fullStr Power Optimal Network-On-Chip Interconnect Design
title_full_unstemmed Power Optimal Network-On-Chip Interconnect Design
title_sort power optimal network-on-chip interconnect design
publishDate 2011
url http://etd.iisc.ernet.in/handle/2005/1408
http://etd.ncsi.iisc.ernet.in/abstracts/1817/G23700-Abs.pdf
work_keys_str_mv AT vikasg poweroptimalnetworkonchipinterconnectdesign
_version_ 1718603128189747200