CrossLight: A Cross-Layer Optimized Silicon Photonic Neural Network Accelerator

Domain-specific neural network accelerators have seen growing interest in recent years due to their improved energy efficiency and performance compared to CPUs and GPUs. In this paper, we propose a novel cross-layer optimized neural network accelerator called CrossLight that leverages silicon photon...

Full description

Saved in:
Bibliographic Details
Published in:2021 58th ACM/IEEE Design Automation Conference (DAC) pp. 1069 - 1074
Main Authors: Sunny, Febin, Mirza, Asif, Nikdast, Mahdi, Pasricha, Sudeep
Format: Conference Proceeding
Language:English
Published: IEEE 05.12.2021
Subjects:
Crosstalk
deep learning
Neural networks
Performance evaluation
Power lasers
Silicon photonics
Thermal engineering
Throughput
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract Domain-specific neural network accelerators have seen growing interest in recent years due to their improved energy efficiency and performance compared to CPUs and GPUs. In this paper, we propose a novel cross-layer optimized neural network accelerator called CrossLight that leverages silicon photonics. CrossLight includes device-level engineering for resilience to process variations and thermal crosstalk, circuit-level tuning enhancements for inference latency reduction, and architecture-level optimizations to enable better resolution, energy-efficiency, and throughput. On average, CrossLight offers 9.5x lower energy-per-bit and 15.9x higher performance-per-watt than state-of-the-art photonic deep learning accelerators.
AbstractList Domain-specific neural network accelerators have seen growing interest in recent years due to their improved energy efficiency and performance compared to CPUs and GPUs. In this paper, we propose a novel cross-layer optimized neural network accelerator called CrossLight that leverages silicon photonics. CrossLight includes device-level engineering for resilience to process variations and thermal crosstalk, circuit-level tuning enhancements for inference latency reduction, and architecture-level optimizations to enable better resolution, energy-efficiency, and throughput. On average, CrossLight offers 9.5x lower energy-per-bit and 15.9x higher performance-per-watt than state-of-the-art photonic deep learning accelerators.
Author Sunny, Febin
Mirza, Asif
Nikdast, Mahdi
Pasricha, Sudeep
Author_xml – sequence: 1
  givenname: Febin
  surname: Sunny
  fullname: Sunny, Febin
  email: febin.sunny@colostate.edu
  organization: Colorado State University,Department of Electrical and Computer Engineering,Fort Collins,CO,USA
– sequence: 2
  givenname: Asif
  surname: Mirza
  fullname: Mirza, Asif
  email: mirza.baig@colostate.edu
  organization: Colorado State University,Department of Electrical and Computer Engineering,Fort Collins,CO,USA
– sequence: 3
  givenname: Mahdi
  surname: Nikdast
  fullname: Nikdast, Mahdi
  email: mahdi.nikdast@colostate.edu
  organization: Colorado State University,Department of Electrical and Computer Engineering,Fort Collins,CO,USA
– sequence: 4
  givenname: Sudeep
  surname: Pasricha
  fullname: Pasricha, Sudeep
  email: sudeep@colostate.edu
  organization: Colorado State University,Department of Electrical and Computer Engineering,Fort Collins,CO,USA
BookMark eNotj8tKxDAYRiMoqGOfQIS8QMfck7or9QrFCup6SJM_TrDTDGlFxqd30Nmcw9l88J2j4zGNgNAVJUtKSXV9WzfUEC2WjDC6rKRRVNEjVFTaUKWk4EwLcoqKaYo9UUQasecZ6pqcpqmNH-v5Btf4r8rW7iDjbjvHTfwBj1_jEF0a8cs6zWmMDj_DV7bDXvN3yp-4dg4GyHZO-QKdBDtMUBy8QO_3d2_NY9l2D09N3ZaWGT2XRgpvTOBM8aqvvHGcBMa5kpo7JwI3QQQDihLdU6Y5F-C9dkqBkF6C7_kCXf7vRgBYbXPc2LxbHW7zXxEgT80
ContentType Conference Proceeding
DBID 6IE
6IH
CBEJK
RIE
RIO
DOI 10.1109/DAC18074.2021.9586161
DatabaseName IEEE Electronic Library (IEL) Conference Proceedings
IEEE Proceedings Order Plan (POP) 1998-present by volume
IEEE Xplore All Conference Proceedings
IEEE Electronic Library (IEL)
IEEE Proceedings Order Plans (POP) 1998-present
DatabaseTitleList
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
EISBN 9781665432740
1665432748
EndPage 1074
ExternalDocumentID 9586161
Genre orig-research
GroupedDBID 6IE
6IH
ACM
ALMA_UNASSIGNED_HOLDINGS
CBEJK
RIE
RIO
ID FETCH-LOGICAL-a287t-854d88f32639b9d8c30f2336573cc4f38f4f8e6107b127334edd7c66e45d5edb3
IEDL.DBID RIE
ISICitedReferencesCount 61
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000766079700179&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
IngestDate Wed Aug 27 02:28:30 EDT 2025
IsPeerReviewed false
IsScholarly true
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-a287t-854d88f32639b9d8c30f2336573cc4f38f4f8e6107b127334edd7c66e45d5edb3
PageCount 6
ParticipantIDs ieee_primary_9586161
PublicationCentury 2000
PublicationDate 2021-Dec.-5
PublicationDateYYYYMMDD 2021-12-05
PublicationDate_xml – month: 12
  year: 2021
  text: 2021-Dec.-5
  day: 05
PublicationDecade 2020
PublicationTitle 2021 58th ACM/IEEE Design Automation Conference (DAC)
PublicationTitleAbbrev DAC
PublicationYear 2021
Publisher IEEE
Publisher_xml – name: IEEE
SSID ssib060584060
Score 2.4672196
Snippet Domain-specific neural network accelerators have seen growing interest in recent years due to their improved energy efficiency and performance compared to CPUs...
SourceID ieee
SourceType Publisher
StartPage 1069
SubjectTerms Crosstalk
deep learning
Neural networks
Performance evaluation
Power lasers
Silicon photonics
Thermal engineering
Throughput
Title CrossLight: A Cross-Layer Optimized Silicon Photonic Neural Network Accelerator
URI https://ieeexplore.ieee.org/document/9586161
WOSCitedRecordID wos000766079700179&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1NS8MwGA5zePCksonf5ODRbGuTNqm3MR0exjbwg91GkzfBwWxldh789b7J6kTw4ilNoQTelDzPk_eLkKtMSwUIpSw3UjAB1jEFkWax48hPhAIVWic8j-R4rGazbNog19tcGGttCD6zHf8YfPlQmrW_KutmiUojr3V2pEw3uVrf_4737iE29eoknaiXdW_7g8iXekERGEed-ttfTVQChgz3_7f6AWn_JOPR6RZmDknDFi0yGXh0G3llfUP7NMzYKEf-TCd4CLwuPi3Qh8US97mg05ey8hVwqa_EkS9xCKHftG8Mgk7ws7fJ0_DucXDP6t4ILEeNUzGVCFDKIfnimc5AGd5zMedpIrkxwnHlhFMWuZHUETIULiyANGlqRQKJBc2PSLMoC3tMqNBcZrHGU8-BAOHjnrRXhU75rhExnJCWN8b8bVP-Yl7b4fTv12dkz9s7RHwk56RZrdb2guyaj2rxvroMe_YFy4qW7g
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1dS8MwFA1jCvqksonf5sFHs61N2qa-jemYWLeBU_Y2mtwEB3OT2fngrzc31ongi0_9gFK4KTnn9H4cQi5SlUhwUMpynQgmwFgmIVAstNzxEyFBeuuEpyzp9-V4nA4r5HLdC2OM8cVnpoGnPpcPC73CX2XNNJJxgFpnA52zym6t768H83sOnVplm07QSpvX7U6Aw16cDAyDRvn0LxsVjyLdnf-9f5fUf9rx6HANNHukYuY1MuggvmWora9om_orluWOQdOB2wZeph8G6MN05lZ6TofPiwJn4FKcxZHP3MEXf9O21g52fKa9Th67N6NOj5XuCCx3KqdgMhIgpXX0i6cqBal5y4acx1HCtRaWSyusNI4dJSpwHIULA5DoODYigsiA4vukOl_MzQGhQvEkDZXb9ywIEFj5pFAXWom-ESEckhoGY_L6NQBjUsbh6O_b52SrN7rPJtlt_-6YbGPsff1HdEKqxXJlTsmmfi-mb8szv36f2xuaNw
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=proceeding&rft.title=2021+58th+ACM%2FIEEE+Design+Automation+Conference+%28DAC%29&rft.atitle=CrossLight%3A+A+Cross-Layer+Optimized+Silicon+Photonic+Neural+Network+Accelerator&rft.au=Sunny%2C+Febin&rft.au=Mirza%2C+Asif&rft.au=Nikdast%2C+Mahdi&rft.au=Pasricha%2C+Sudeep&rft.date=2021-12-05&rft.pub=IEEE&rft.spage=1069&rft.epage=1074&rft_id=info:doi/10.1109%2FDAC18074.2021.9586161&rft.externalDocID=9586161