Energy-efficient and bandwidth-reconfigurable photonic networks for high-performance computing (HPC) systems

Avinash Karanth Kodi, Ahmed Louri

Research output: Contribution to journalArticle

36 Citations (Scopus)

Abstract

Optical interconnects are becoming ubiquitous for short-range communication within boards and racks due to higher communication bandwidth at lower power dissipation when compared to metallic interconnects. Efficient multiplexing techniques (wavelengths, time, and space) allow bandwidths to scale; static or predetermined resource allocation of wavelengths can be detrimental to network performance for nonuniform (adversial) workloads. Dynamic bandwidth reallocation (DBR) based on actual traffic pattern can lead to improved network performance by utilizing idle resources. While DBR techniques can alleviate interconnection bottlenecks, power consumption also increases considerably with increase in bit rate and channels. In this paper, we propose to improve the performance of optical interconnects using DBR techniques and simultaneously optimize the power consumption using dynamic power management (DPM) techniques. DBR reallocates idle channels to busy channels (wavelengths) for improving throughput, and DPM regulates the bit rates and supply voltages for the individual channels. A reconfigurable optoelectronic architecture and a performance adaptive algorithm for implementing DBR and DPM are proposed in this paper. Our proposed reconfiguration algorithm achieves a significant reduction in power consumption and considerable improvement in throughput, with a marginal increase in latency for synthetic and real (Splash-2) traffic traces.

Original languageEnglish (US)
Article number5510030
Pages (from-to)384-395
Number of pages12
JournalIEEE Journal on Selected Topics in Quantum Electronics
Volume17
Issue number2
DOIs
StatePublished - Mar 2011

Fingerprint

Photonics
photonics
bandwidth
Bandwidth
Electric power utilization
optical interconnects
Optical interconnects
energy
Network performance
Wavelength
traffic
communication
Throughput
wavelengths
resource allocation
racks
Communication
Adaptive algorithms
multiplexing
Multiplexing

Keywords

  • High-performance computing (HPC)
  • performance modeling
  • power aware
  • reconfigurable optical interconnects

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Atomic and Molecular Physics, and Optics

Cite this

Energy-efficient and bandwidth-reconfigurable photonic networks for high-performance computing (HPC) systems. / Kodi, Avinash Karanth; Louri, Ahmed.

In: IEEE Journal on Selected Topics in Quantum Electronics, Vol. 17, No. 2, 5510030, 03.2011, p. 384-395.

Research output: Contribution to journalArticle

@article{55fc876bba2f4bfb86b08ad12e04d146,
title = "Energy-efficient and bandwidth-reconfigurable photonic networks for high-performance computing (HPC) systems",
abstract = "Optical interconnects are becoming ubiquitous for short-range communication within boards and racks due to higher communication bandwidth at lower power dissipation when compared to metallic interconnects. Efficient multiplexing techniques (wavelengths, time, and space) allow bandwidths to scale; static or predetermined resource allocation of wavelengths can be detrimental to network performance for nonuniform (adversial) workloads. Dynamic bandwidth reallocation (DBR) based on actual traffic pattern can lead to improved network performance by utilizing idle resources. While DBR techniques can alleviate interconnection bottlenecks, power consumption also increases considerably with increase in bit rate and channels. In this paper, we propose to improve the performance of optical interconnects using DBR techniques and simultaneously optimize the power consumption using dynamic power management (DPM) techniques. DBR reallocates idle channels to busy channels (wavelengths) for improving throughput, and DPM regulates the bit rates and supply voltages for the individual channels. A reconfigurable optoelectronic architecture and a performance adaptive algorithm for implementing DBR and DPM are proposed in this paper. Our proposed reconfiguration algorithm achieves a significant reduction in power consumption and considerable improvement in throughput, with a marginal increase in latency for synthetic and real (Splash-2) traffic traces.",
keywords = "High-performance computing (HPC), performance modeling, power aware, reconfigurable optical interconnects",
author = "Kodi, {Avinash Karanth} and Ahmed Louri",
year = "2011",
month = "3",
doi = "10.1109/JSTQE.2010.2051419",
language = "English (US)",
volume = "17",
pages = "384--395",
journal = "IEEE Journal of Selected Topics in Quantum Electronics",
issn = "1077-260X",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "2",

}

TY - JOUR

T1 - Energy-efficient and bandwidth-reconfigurable photonic networks for high-performance computing (HPC) systems

AU - Kodi, Avinash Karanth

AU - Louri, Ahmed

PY - 2011/3

Y1 - 2011/3

N2 - Optical interconnects are becoming ubiquitous for short-range communication within boards and racks due to higher communication bandwidth at lower power dissipation when compared to metallic interconnects. Efficient multiplexing techniques (wavelengths, time, and space) allow bandwidths to scale; static or predetermined resource allocation of wavelengths can be detrimental to network performance for nonuniform (adversial) workloads. Dynamic bandwidth reallocation (DBR) based on actual traffic pattern can lead to improved network performance by utilizing idle resources. While DBR techniques can alleviate interconnection bottlenecks, power consumption also increases considerably with increase in bit rate and channels. In this paper, we propose to improve the performance of optical interconnects using DBR techniques and simultaneously optimize the power consumption using dynamic power management (DPM) techniques. DBR reallocates idle channels to busy channels (wavelengths) for improving throughput, and DPM regulates the bit rates and supply voltages for the individual channels. A reconfigurable optoelectronic architecture and a performance adaptive algorithm for implementing DBR and DPM are proposed in this paper. Our proposed reconfiguration algorithm achieves a significant reduction in power consumption and considerable improvement in throughput, with a marginal increase in latency for synthetic and real (Splash-2) traffic traces.

AB - Optical interconnects are becoming ubiquitous for short-range communication within boards and racks due to higher communication bandwidth at lower power dissipation when compared to metallic interconnects. Efficient multiplexing techniques (wavelengths, time, and space) allow bandwidths to scale; static or predetermined resource allocation of wavelengths can be detrimental to network performance for nonuniform (adversial) workloads. Dynamic bandwidth reallocation (DBR) based on actual traffic pattern can lead to improved network performance by utilizing idle resources. While DBR techniques can alleviate interconnection bottlenecks, power consumption also increases considerably with increase in bit rate and channels. In this paper, we propose to improve the performance of optical interconnects using DBR techniques and simultaneously optimize the power consumption using dynamic power management (DPM) techniques. DBR reallocates idle channels to busy channels (wavelengths) for improving throughput, and DPM regulates the bit rates and supply voltages for the individual channels. A reconfigurable optoelectronic architecture and a performance adaptive algorithm for implementing DBR and DPM are proposed in this paper. Our proposed reconfiguration algorithm achieves a significant reduction in power consumption and considerable improvement in throughput, with a marginal increase in latency for synthetic and real (Splash-2) traffic traces.

KW - High-performance computing (HPC)

KW - performance modeling

KW - power aware

KW - reconfigurable optical interconnects

UR - http://www.scopus.com/inward/record.url?scp=79954581808&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79954581808&partnerID=8YFLogxK

U2 - 10.1109/JSTQE.2010.2051419

DO - 10.1109/JSTQE.2010.2051419

M3 - Article

AN - SCOPUS:79954581808

VL - 17

SP - 384

EP - 395

JO - IEEE Journal of Selected Topics in Quantum Electronics

JF - IEEE Journal of Selected Topics in Quantum Electronics

SN - 1077-260X

IS - 2

M1 - 5510030

ER -