LoraX: Loss-aware approximations for energy-efficient silicon photonic networks-on-chip

Febin Sunny; Asif Mirza; Ishan Thakkar; Sudeep Pasricha; Mahdi Nikdast

doi:10.1145/3386263.3406919

LoraX: Loss-aware approximations for energy-efficient silicon photonic networks-on-chip

Febin Sunny, Asif Mirza, Ishan Thakkar, Sudeep Pasricha, Mahdi Nikdast

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

9 Scopus citations

Abstract

The approximate computing paradigm advocates for relaxing accuracy goals in applications to improve energy-efficiency and performance. Recently, this paradigm has been explored to improve the energy efficiency of silicon photonic networks-on-chip (PNoCs). In this paper, we propose a novel framework (LORAX) to enable more aggressive approximation during communication over silicon photonic links in PNoCs. This is the first work that considers loss-aware laser power management and multilevel signaling to enable effective data approximation and energy-efficiency in PNoCs. Simulation results show that our framework can achieve up to 31.4% lower laser power consumption and up to 12.2% better energy efficiency than the best known prior work on approximate communication in PNoCs, for the same application output quality.

Original language	English
Title of host publication	GLSVLSI 2020 - Proceedings of the 2020 Great Lakes Symposium on VLSI
Pages	235-240
Number of pages	6
ISBN (Electronic)	9781450379441
DOIs	https://doi.org/10.1145/3386263.3406919
State	Published - Sep 7 2020
Event	30th Great Lakes Symposium on VLSI, GLSVLSI 2020 - Virtual, Online, China Duration: Sep 7 2020 → Sep 9 2020

Publication series

Name	Proceedings of the ACM Great Lakes Symposium on VLSI, GLSVLSI

Conference

Conference	30th Great Lakes Symposium on VLSI, GLSVLSI 2020
Country/Territory	China
City	Virtual, Online
Period	9/7/20 → 9/9/20

Bibliographical note

Publisher Copyright:
© 2020 Association for Computing Machinery.

Funding

This research was supported by the National Science Foundation (NSF) under grant numbers CCF-1813370 and CCF-2006788.

Funders	Funder number
National Science Foundation Arctic Social Science Program	CCF-2006788, CCF-1813370
National Science Foundation Arctic Social Science Program
Directorate for Computer and Information Science and Engineering	1813370
Directorate for Computer and Information Science and Engineering

Keywords

Approximate computing
Energy efficiency
PNoCs

ASJC Scopus subject areas

General Engineering

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1145/3386263.3406919

Cite this

@inproceedings{316c0b036a614b3c81d0aa1748659ee0,

title = "LoraX: Loss-aware approximations for energy-efficient silicon photonic networks-on-chip",

abstract = "The approximate computing paradigm advocates for relaxing accuracy goals in applications to improve energy-efficiency and performance. Recently, this paradigm has been explored to improve the energy efficiency of silicon photonic networks-on-chip (PNoCs). In this paper, we propose a novel framework (LORAX) to enable more aggressive approximation during communication over silicon photonic links in PNoCs. This is the first work that considers loss-aware laser power management and multilevel signaling to enable effective data approximation and energy-efficiency in PNoCs. Simulation results show that our framework can achieve up to 31.4% lower laser power consumption and up to 12.2% better energy efficiency than the best known prior work on approximate communication in PNoCs, for the same application output quality.",

keywords = "Approximate computing, Energy efficiency, PNoCs",

author = "Febin Sunny and Asif Mirza and Ishan Thakkar and Sudeep Pasricha and Mahdi Nikdast",

note = "Publisher Copyright: {\textcopyright} 2020 Association for Computing Machinery.; 30th Great Lakes Symposium on VLSI, GLSVLSI 2020 ; Conference date: 07-09-2020 Through 09-09-2020",

year = "2020",

month = sep,

day = "7",

doi = "10.1145/3386263.3406919",

language = "English",

series = "Proceedings of the ACM Great Lakes Symposium on VLSI, GLSVLSI",

pages = "235--240",

booktitle = "GLSVLSI 2020 - Proceedings of the 2020 Great Lakes Symposium on VLSI",

}

Sunny, F, Mirza, A, Thakkar, I, Pasricha, S & Nikdast, M 2020, LoraX: Loss-aware approximations for energy-efficient silicon photonic networks-on-chip. in GLSVLSI 2020 - Proceedings of the 2020 Great Lakes Symposium on VLSI. Proceedings of the ACM Great Lakes Symposium on VLSI, GLSVLSI, pp. 235-240, 30th Great Lakes Symposium on VLSI, GLSVLSI 2020, Virtual, Online, China, 9/7/20. https://doi.org/10.1145/3386263.3406919

LoraX: Loss-aware approximations for energy-efficient silicon photonic networks-on-chip. / Sunny, Febin; Mirza, Asif; Thakkar, Ishan et al.
GLSVLSI 2020 - Proceedings of the 2020 Great Lakes Symposium on VLSI. 2020. p. 235-240 (Proceedings of the ACM Great Lakes Symposium on VLSI, GLSVLSI).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - LoraX

T2 - 30th Great Lakes Symposium on VLSI, GLSVLSI 2020

AU - Sunny, Febin

AU - Mirza, Asif

AU - Thakkar, Ishan

AU - Pasricha, Sudeep

AU - Nikdast, Mahdi

PY - 2020/9/7

Y1 - 2020/9/7

N2 - The approximate computing paradigm advocates for relaxing accuracy goals in applications to improve energy-efficiency and performance. Recently, this paradigm has been explored to improve the energy efficiency of silicon photonic networks-on-chip (PNoCs). In this paper, we propose a novel framework (LORAX) to enable more aggressive approximation during communication over silicon photonic links in PNoCs. This is the first work that considers loss-aware laser power management and multilevel signaling to enable effective data approximation and energy-efficiency in PNoCs. Simulation results show that our framework can achieve up to 31.4% lower laser power consumption and up to 12.2% better energy efficiency than the best known prior work on approximate communication in PNoCs, for the same application output quality.

AB - The approximate computing paradigm advocates for relaxing accuracy goals in applications to improve energy-efficiency and performance. Recently, this paradigm has been explored to improve the energy efficiency of silicon photonic networks-on-chip (PNoCs). In this paper, we propose a novel framework (LORAX) to enable more aggressive approximation during communication over silicon photonic links in PNoCs. This is the first work that considers loss-aware laser power management and multilevel signaling to enable effective data approximation and energy-efficiency in PNoCs. Simulation results show that our framework can achieve up to 31.4% lower laser power consumption and up to 12.2% better energy efficiency than the best known prior work on approximate communication in PNoCs, for the same application output quality.

KW - Approximate computing

KW - Energy efficiency

KW - PNoCs

UR - http://www.scopus.com/inward/record.url?scp=85090926042&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85090926042&partnerID=8YFLogxK

U2 - 10.1145/3386263.3406919

DO - 10.1145/3386263.3406919

M3 - Conference contribution

AN - SCOPUS:85090926042

T3 - Proceedings of the ACM Great Lakes Symposium on VLSI, GLSVLSI

SP - 235

EP - 240

BT - GLSVLSI 2020 - Proceedings of the 2020 Great Lakes Symposium on VLSI

Y2 - 7 September 2020 through 9 September 2020

ER -

LoraX: Loss-aware approximations for energy-efficient silicon photonic networks-on-chip

Abstract

Publication series

Conference

Bibliographical note

Funding

Keywords

ASJC Scopus subject areas

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this