Device Variation Effects on Neural Network Inference Accuracy in Analog In‐Memory Computing Systems

In analog in‐memory computing systems based on nonvolatile memories such as resistive random‐access memory (RRAM), neural network models are often trained offline and then the weights are programmed onto memory devices as conductance values. The programmed weight values inevitably deviate from the t...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Advanced intelligent systems Ročník 4; číslo 8
Hlavní autori: Wang, Qiwen, Park, Yongmo, Lu, Wei D.
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Weinheim John Wiley & Sons, Inc 01.08.2022
Wiley
Predmet:
Accuracy
analog computing
Analog to digital converters
Arrays
Circuits
Computation
Datasets
deep neural networks
emerging memory
Energy efficiency
in-memory computing
Inference
Internet of Things
Memory devices
Neural networks
process-in-memory
Programming
RRAM
Training
Workloads
ISSN:2640-4567, 2640-4567
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Abstract In analog in‐memory computing systems based on nonvolatile memories such as resistive random‐access memory (RRAM), neural network models are often trained offline and then the weights are programmed onto memory devices as conductance values. The programmed weight values inevitably deviate from the target values during the programming process. This effect can be pronounced for emerging memories such as RRAM, PcRAM, and MRAM due to the stochastic nature during programming. Unlike noise, these weight deviations do not change during inference. The performance of neural network models is investigated against this programming variation under realistic system limitations, including limited device on/off ratios, memory array size, analog‐to‐digital converter (ADC) characteristics, and signed weight representations. Approaches to mitigate such device and circuit nonidealities through architecture‐aware training are also evaluated. The effectiveness of variation injection during training to improve the inference robustness, as well as the effects of different neural network training parameters such as learning rate schedule, will be discussed. In nonvolatile memory‐based analog in‐memory computing systems, variations during device programming can cause neural‐network inference accuracy to degrade since the stored weights will differ from those in the original models. Herein, the performance of deep neural‐network models is investigated against this effect under realistic system limitations, including limited device on/off ratios, memory array size, circuit characteristics, and signed weight representations.
AbstractList In analog in‐memory computing systems based on nonvolatile memories such as resistive random‐access memory (RRAM), neural network models are often trained offline and then the weights are programmed onto memory devices as conductance values. The programmed weight values inevitably deviate from the target values during the programming process. This effect can be pronounced for emerging memories such as RRAM, PcRAM, and MRAM due to the stochastic nature during programming. Unlike noise, these weight deviations do not change during inference. The performance of neural network models is investigated against this programming variation under realistic system limitations, including limited device on/off ratios, memory array size, analog‐to‐digital converter (ADC) characteristics, and signed weight representations. Approaches to mitigate such device and circuit nonidealities through architecture‐aware training are also evaluated. The effectiveness of variation injection during training to improve the inference robustness, as well as the effects of different neural network training parameters such as learning rate schedule, will be discussed. In nonvolatile memory‐based analog in‐memory computing systems, variations during device programming can cause neural‐network inference accuracy to degrade since the stored weights will differ from those in the original models. Herein, the performance of deep neural‐network models is investigated against this effect under realistic system limitations, including limited device on/off ratios, memory array size, circuit characteristics, and signed weight representations.
In analog in‐memory computing systems based on nonvolatile memories such as resistive random‐access memory (RRAM), neural network models are often trained offline and then the weights are programmed onto memory devices as conductance values. The programmed weight values inevitably deviate from the target values during the programming process. This effect can be pronounced for emerging memories such as RRAM, PcRAM, and MRAM due to the stochastic nature during programming. Unlike noise, these weight deviations do not change during inference. The performance of neural network models is investigated against this programming variation under realistic system limitations, including limited device on/off ratios, memory array size, analog‐to‐digital converter (ADC) characteristics, and signed weight representations. Approaches to mitigate such device and circuit nonidealities through architecture‐aware training are also evaluated. The effectiveness of variation injection during training to improve the inference robustness, as well as the effects of different neural network training parameters such as learning rate schedule, will be discussed.
Author Wang, Qiwen
Lu, Wei D.
Park, Yongmo
Author_xml – sequence: 1
  givenname: Qiwen
  surname: Wang
  fullname: Wang, Qiwen
  organization: University of Michigan
– sequence: 2
  givenname: Yongmo
  surname: Park
  fullname: Park, Yongmo
  organization: University of Michigan
– sequence: 3
  givenname: Wei D.
  orcidid: 0000-0003-4731-1976
  surname: Lu
  fullname: Lu, Wei D.
  email: wluee@umich.edu
  organization: University of Michigan
BookMark eNqFUUtPGzEQtiqQSoFrzytxTvBjn8copSUSbQ88pJ6syew4ctisU9sp2hs_gd_IL8EhiKJKFacZe77H2N8ntte7nhj7LPhYcC5PwYZhLLlMB9E0H9iBLHM-youy2nvTf2THISx5IohKcFkdMPpCfyxSdgPeQrSuz86MIYwhS-0P2njoUol3zt9ms96Qpz6hJ4hpgkNm-2zSQ-cWafh4__CdVs4P2dSt1pto-0V2OYRIq3DE9g10gY5f6iG7_np2NT0fXfz8NptOLkaoCtWMapUXbZVTjdyYEjiailNatt2ua5oKUM2RWsEbqNuWlCiNrDjWoqEcJKE6ZLOdbutgqdfersAP2oHVzxfOLzT4aLEjrQouymoOwog2rwXMFVKJhTESWoKiSFonO621d783FKJeuo1Pjw06mapScdnIhBrvUOhdCJ7Mq6vgepuM3iajX5NJhPwfAtr4_PPRg-3-T2t2tDvb0fCOiZ7MLn_95T4BXT2m4Q
CitedBy_id crossref_primary_10_3390_nano12203582
crossref_primary_10_1002_aelm_202201006
crossref_primary_10_1109_JETCAS_2022_3227471
crossref_primary_10_1002_adma_202405145
crossref_primary_10_1002_advs_202308460
crossref_primary_10_1002_aisy_202200127
crossref_primary_10_1002_aelm_202400106
crossref_primary_10_1109_LED_2022_3192262
crossref_primary_10_1002_aisy_202300763
crossref_primary_10_1002_apxr_202300085
crossref_primary_10_1088_1674_4926_45_1_012301
crossref_primary_10_1002_adma_202305465
crossref_primary_10_1016_j_microrel_2025_115594
Cites_doi 10.1109/TVLSI.2019.2893256
10.1109/CICC.2017.7993628
10.1109/IJCNN.2019.8851966
10.1109/ISSCC.2011.5746281
10.1038/s41586-020-1942-4
10.5244/C.30.87
10.1145/3316781.3317770
10.1109/SiPS50750.2020.9195245
10.1038/s41928-019-0270-x
10.1109/IEDM19573.2019.8993641
10.1016/B978-0-08-102782-0.00009-5
10.1109/IEDM19573.2019.8993491
10.7873/DATE.2013.266
10.1109/JSSC.2017.2712626
10.1109/IEDM.2016.7838434
10.1109/ISCAS.2019.8702245
10.1038/s41928-018-0100-6
10.1109/TCSI.2021.3079980
10.1038/s41598-020-62676-7
10.1162/neco.1996.8.3.643
10.1109/AICAS48895.2020.9073942
10.1109/TNANO.2017.2784364
10.1109/MCAS.2021.3092533
10.1038/s41467-020-16108-9
10.1109/72.238328
10.1109/WACV.2017.58
ContentType Journal Article
Copyright 2022 The Authors. Advanced Intelligent Systems published by Wiley‐VCH GmbH
2022. This work is published under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Copyright_xml – notice: 2022 The Authors. Advanced Intelligent Systems published by Wiley‐VCH GmbH
– notice: 2022. This work is published under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
DBID 24P
AAYXX
CITATION
8FE
8FG
ABUWG
AFKRA
ARAPS
AZQEC
BENPR
BGLVJ
CCPQU
DWQXO
HCIFZ
P5Z
P62
PHGZM
PHGZT
PIMPY
PKEHL
PQEST
PQGLB
PQQKQ
PQUKI
PRINS
DOA
DOI 10.1002/aisy.202100199
DatabaseName Wiley Online Library Open Access
CrossRef
ProQuest SciTech Collection
ProQuest Technology Collection
ProQuest Central
ProQuest Central UK/Ireland
Advanced Technologies & Computer Science Collection
ProQuest Central Essentials
ProQuest Central
ProQuest Technology Collection
ProQuest One
ProQuest Central Korea
SciTech Premium Collection (via ProQuest)
Advanced Technologies & Aerospace Database
ProQuest Advanced Technologies & Aerospace Collection
ProQuest Central Premium
ProQuest One Academic
ProQuest Publicly Available Content Database
ProQuest One Academic Middle East (New)
ProQuest One Academic Eastern Edition (DO NOT USE)
ProQuest One Applied & Life Sciences
ProQuest One Academic (retired)
ProQuest One Academic UKI Edition
ProQuest Central China
DOAJ Directory of Open Access Journals
DatabaseTitle CrossRef
Publicly Available Content Database
Advanced Technologies & Aerospace Collection
Technology Collection
ProQuest One Academic Middle East (New)
ProQuest Advanced Technologies & Aerospace Collection
ProQuest Central Essentials
ProQuest One Academic Eastern Edition
ProQuest Central (Alumni Edition)
SciTech Premium Collection
ProQuest One Community College
ProQuest Technology Collection
ProQuest SciTech Collection
ProQuest Central China
ProQuest Central
Advanced Technologies & Aerospace Database
ProQuest One Applied & Life Sciences
ProQuest One Academic UKI Edition
ProQuest Central Korea
ProQuest Central (New)
ProQuest One Academic
ProQuest One Academic (New)
DatabaseTitleList
Publicly Available Content Database
Database_xml – sequence: 1
  dbid: DOA
  name: DOAJ: Directory of Open Access Journal (DOAJ)
  url: https://www.doaj.org/
  sourceTypes: Open Website
– sequence: 2
  dbid: 24P
  name: Wiley Open Access
  url: https://authorservices.wiley.com/open-science/open-access/browse-journals.html
  sourceTypes: Publisher
– sequence: 3
  dbid: PIMPY
  name: Publicly Available Content Database
  url: http://search.proquest.com/publiccontent
  sourceTypes: Aggregation Database
DeliveryMethod fulltext_linktorsrc
EISSN 2640-4567
EndPage n/a
ExternalDocumentID oai_doaj_org_article_350167ba1f1d481ab3ce6c5ff2adea55
10_1002_aisy_202100199
AISY202100199
Genre article
GrantInformation_xml – fundername: National Science Foundation
  funderid: CCF-1900675
GroupedDBID 0R~
1OC
24P
AAFWJ
AAHHS
ACCFJ
ACCMX
ACXQS
ADKYN
ADZMN
ADZOD
AEEZP
AEQDE
AFKRA
AIWBW
AJBDE
ALMA_UNASSIGNED_HOLDINGS
ALUQN
ARAPS
ARCSS
AVUZU
BENPR
BGLVJ
CCPQU
EBS
EJD
GROUPED_DOAJ
HCIFZ
IAO
ITC
M~E
OK1
PIMPY
WIN
AAMMB
AAYXX
ADMLS
AEFGJ
AFFHD
AFPKN
AGXDD
AIDQK
AIDYY
CITATION
ICD
PHGZM
PHGZT
PQGLB
8FE
8FG
ABUWG
AZQEC
DWQXO
P62
PKEHL
PQEST
PQQKQ
PQUKI
PRINS
ID FETCH-LOGICAL-c3539-8345d74e8c0ff6a0cf70e002d2171f97ac3bced109a8dde316f270c819e4a2ec3
IEDL.DBID P5Z
ISICitedReferencesCount 18
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000743074700001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 2640-4567
IngestDate Fri Oct 03 12:51:00 EDT 2025
Sun Nov 09 08:18:06 EST 2025
Sat Nov 29 07:23:11 EST 2025
Tue Nov 18 21:51:56 EST 2025
Wed Jan 22 16:22:56 EST 2025
IsDoiOpenAccess true
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Issue 8
Language English
License Attribution
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c3539-8345d74e8c0ff6a0cf70e002d2171f97ac3bced109a8dde316f270c819e4a2ec3
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ORCID 0000-0003-4731-1976
OpenAccessLink https://www.proquest.com/docview/2703630292?pq-origsite=%requestingapplication%
PQID 2703630292
PQPubID 5064933
PageCount 12
ParticipantIDs doaj_primary_oai_doaj_org_article_350167ba1f1d481ab3ce6c5ff2adea55
proquest_journals_2703630292
crossref_primary_10_1002_aisy_202100199
crossref_citationtrail_10_1002_aisy_202100199
wiley_primary_10_1002_aisy_202100199_AISY202100199
PublicationCentury 2000
PublicationDate August 2022
2022-08-00
20220801
2022-08-01
PublicationDateYYYYMMDD 2022-08-01
PublicationDate_xml – month: 08
  year: 2022
  text: August 2022
PublicationDecade 2020
PublicationPlace Weinheim
PublicationPlace_xml – name: Weinheim
PublicationTitle Advanced intelligent systems
PublicationYear 2022
Publisher John Wiley & Sons, Inc
Wiley
Publisher_xml – name: John Wiley & Sons, Inc
– name: Wiley
References 2021; 68
2018; 17
2017; 52
2021; 21
2011
2019; 2
2021
2018; 1
2020
2019
2020; 577
2019; 27
2018
2017
2016
2015
2020; 11
1992
2013
1996; 8
1993; 4
e_1_2_11_10_1
e_1_2_11_32_1
e_1_2_11_31_1
e_1_2_11_30_1
e_1_2_11_36_1
e_1_2_11_14_1
e_1_2_11_13_1
e_1_2_11_35_1
e_1_2_11_12_1
e_1_2_11_34_1
e_1_2_11_11_1
e_1_2_11_33_1
e_1_2_11_7_1
e_1_2_11_6_1
e_1_2_11_28_1
e_1_2_11_5_1
e_1_2_11_27_1
Noh H. (e_1_2_11_29_1) 2017
e_1_2_11_4_1
e_1_2_11_26_1
e_1_2_11_3_1
e_1_2_11_2_1
e_1_2_11_21_1
e_1_2_11_20_1
e_1_2_11_25_1
e_1_2_11_24_1
e_1_2_11_9_1
e_1_2_11_23_1
e_1_2_11_8_1
e_1_2_11_22_1
e_1_2_11_18_1
e_1_2_11_17_1
e_1_2_11_16_1
e_1_2_11_15_1
e_1_2_11_37_1
e_1_2_11_38_1
e_1_2_11_19_1
References_xml – start-page: 1393
  year: 2020
  end-page: 13696
– start-page: 1
  year: 2019
  end-page: 8
– volume: 17
  start-page: 184
  year: 2018
  publication-title: IEEE Trans. Nanotechnol.
– start-page: 200
  year: 2011
  end-page: 202
– start-page: 221
  year: 2020
  end-page: 254
– volume: 2
  start-page: 290
  year: 2019
  publication-title: Nat. Electron.
– start-page: 1
  year: 2019
  end-page: 6
– start-page: 141
  year: 2020
  end-page: 144
– volume: 577
  start-page: 641
  year: 2020
  publication-title: Nature
– year: 2021
– start-page: 329
  year: 2011
  end-page: 332
– volume: 4
  start-page: 722
  year: 1993
  publication-title: IEEE Trans. Neural Netw.
– volume: 68
  start-page: 4470
  year: 2021
  publication-title: IEEE Trans. Circuits Syst. I. Regul. Pap.
– year: 2018
– volume: 27
  start-page: 1455
  year: 2019
  publication-title: IEEE Trans. Very Large Scale Integr. Syst.
– volume: 11
  start-page: 2473
  year: 2020
  publication-title: Nat. Commun.
– start-page: 87.1
  year: 2016
  end-page: 87.12
– start-page: 1
  year: 2021
  end-page: 5
– start-page: 5110
  year: 2017
  end-page: 5119
– start-page: 1
  year: 2017
  end-page: 4
– year: 2020
– start-page: 14.4.1
  year: 2019
  end-page: 14.4.4
– start-page: 1285
  year: 2013
  end-page: 1290
– start-page: 16.7.1
  year: 2017
– volume: 52
  start-page: 2679
  year: 2017
  publication-title: IEEE J. Solid-State Circuits
– start-page: 769
  year: 1992
  end-page: 774
– start-page: 2575
  year: 2015
  end-page: 2583
– start-page: 464
  year: 2017
  end-page: 472
– volume: 21
  start-page: 31
  year: 2021
  publication-title: IEEE Circuits Syst. Mag.
– volume: 8
  start-page: 643
  year: 1996
  publication-title: Neural Comput.
– start-page: 1
  year: 2020
  end-page: 6
– start-page: 132
  year: 2011
  end-page: 137
– volume: 1
  start-page: 411
  year: 2018
  publication-title: Nat. Electron.
– year: 2019
– start-page: 32.5.1
  year: 2019
  end-page: 32.5.4
– year: 2015
– ident: e_1_2_11_30_1
  doi: 10.1109/TVLSI.2019.2893256
– ident: e_1_2_11_19_1
– ident: e_1_2_11_23_1
  doi: 10.1109/CICC.2017.7993628
– ident: e_1_2_11_32_1
  doi: 10.1109/IJCNN.2019.8851966
– ident: e_1_2_11_37_1
  doi: 10.1109/ISSCC.2011.5746281
– ident: e_1_2_11_2_1
– ident: e_1_2_11_8_1
  doi: 10.1038/s41586-020-1942-4
– ident: e_1_2_11_20_1
  doi: 10.5244/C.30.87
– ident: e_1_2_11_21_1
– ident: e_1_2_11_33_1
  doi: 10.1145/3316781.3317770
– ident: e_1_2_11_34_1
  doi: 10.1109/SiPS50750.2020.9195245
– ident: e_1_2_11_7_1
  doi: 10.1038/s41928-019-0270-x
– ident: e_1_2_11_13_1
  doi: 10.1109/IEDM19573.2019.8993641
– ident: e_1_2_11_4_1
  doi: 10.1016/B978-0-08-102782-0.00009-5
– ident: e_1_2_11_5_1
  doi: 10.1109/IEDM19573.2019.8993491
– ident: e_1_2_11_38_1
– ident: e_1_2_11_24_1
  doi: 10.7873/DATE.2013.266
– start-page: 5110
  volume-title: Advances in Neural Information Processing Systems
  year: 2017
  ident: e_1_2_11_29_1
– ident: e_1_2_11_31_1
  doi: 10.1109/JSSC.2017.2712626
– ident: e_1_2_11_14_1
  doi: 10.1109/IEDM.2016.7838434
– ident: e_1_2_11_10_1
  doi: 10.1109/ISCAS.2019.8702245
– ident: e_1_2_11_15_1
– ident: e_1_2_11_22_1
  doi: 10.1038/s41928-018-0100-6
– ident: e_1_2_11_26_1
– ident: e_1_2_11_36_1
  doi: 10.1109/TCSI.2021.3079980
– ident: e_1_2_11_18_1
– ident: e_1_2_11_28_1
– ident: e_1_2_11_9_1
  doi: 10.1038/s41598-020-62676-7
– ident: e_1_2_11_11_1
– ident: e_1_2_11_27_1
  doi: 10.1162/neco.1996.8.3.643
– ident: e_1_2_11_16_1
  doi: 10.1109/AICAS48895.2020.9073942
– ident: e_1_2_11_12_1
  doi: 10.1109/TNANO.2017.2784364
– ident: e_1_2_11_3_1
  doi: 10.1109/MCAS.2021.3092533
– ident: e_1_2_11_6_1
  doi: 10.1038/s41467-020-16108-9
– ident: e_1_2_11_17_1
– ident: e_1_2_11_25_1
  doi: 10.1109/72.238328
– ident: e_1_2_11_35_1
  doi: 10.1109/WACV.2017.58
SSID ssj0002171027
Score 2.300955
Snippet In analog in‐memory computing systems based on nonvolatile memories such as resistive random‐access memory (RRAM), neural network models are often trained...
SourceID doaj
proquest
crossref
wiley
SourceType Open Website
Aggregation Database
Enrichment Source
Index Database
Publisher
SubjectTerms Accuracy
analog computing
Analog to digital converters
Arrays
Circuits
Computation
Datasets
deep neural networks
emerging memory
Energy efficiency
in-memory computing
Inference
Internet of Things
Memory devices
Neural networks
process-in-memory
Programming
RRAM
Training
Workloads
SummonAdditionalLinks – databaseName: DOAJ Directory of Open Access Journals
  dbid: DOA
  link: http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrZ1LS8QwEICDiAcvoqhYXSUHwVMxTdomPa4vFHQRfKCnkE4TWZBVdlXYmz_B3-gvcZI-XA_ixVvbpCGZTDITknxDyG4pM8AFWxFzMFmcOmbj0mYqttxUBlJeyhAO6PZcDgbq7q64nAn15c-E1XjgWnD7fuMrl6VJXFKlKjGlAJtD5hwWZk0W6KVMFjOLKT8Ho6ONllO2lEbG981wMsXlIPfMoQB6_bZCAdb_w8Oc9VODoTlZJkuNh0j7dc1WyJwdrRJ7ZP2Qpre4tA2ypDV2eELx0QM28I9BfaKbnrV3-GgfAFNgSocj6ukjTw-Y-Pn-ceGP105pHdEBbRdtuOVr5Obk-PrwNG4iJMQgMlHESqRZJVOrgDmXGwZOMoutrXz7XSENiBJslbDCKJzHRJI7LhmgF2BTwy2IdTI_ehrZDUIrHMhKYBlhi9fykpvcCPROnIJUKBeRuJWYhgYf7qNYPOoafMy1l7DuJByRvS7_cw3O-DXnge-ALpcHXocPqAa6UQP9lxpEpNd2n25G4UTzQBdjvOAR4aFL_6iK7p9d3Xdvm_9RsS2yyP2NiXBmsEfmX8avdpsswNvLcDLeCVr7Bc7A8fc
  priority: 102
  providerName: Directory of Open Access Journals
– databaseName: Wiley Online Library Open Access
  dbid: 24P
  link: http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV1LT9wwELYK7YELUJWKhQX5UKknC8dOYue4vFSkdrVSAdGT5UxstBLaRRtA2hs_gd_IL2HsZFP2gCrRWxw_ZI3n5cd8Q8i3UmWAG7aCCbAZSz13rHSZZk7YykIqShXTAV3-VMOhvroqRq-i-Bt8iO7ALUhG1NdBwG1ZH_wFDbXjeo77OxFAhIpihXxMEqkCX4t01J2yoMONFjTETKPh5wy9BbVAbuTiYHmIJcsUAfyXvM7Xvms0Pqcb_z_tTbLeOp500HDKZ_LBTb4Qd-yCpqCXuGOOS0QbNOOa4mfA7cAew-ahOD1bhAbSAQDWwJyOJzSAmkyvsfL58elXeLU7p02iCDSJtIVD3yIXpyfnRz9Ym3iBgcxkwbRMs0qlTgP3PrccvOIOp18FcvpCWZAluCrhhdWoHmWSe6E4oHPhUiscyK9kdTKduG1CK9QPWuIY8ebYiVLY3Ep0eryGVGrfI2xBdAMtKnlIjnFjGjxlYQLBTEewHvnetb9t8DjebHkY1rBrFXC044_p7Nq0YmnCtWquSpv4pEp1YksJLofMe2RVZ7OsR_oLDjCtcNdGRNAyLgrRIyKu9T-mYgZnv_90pZ33dNolayIEXsSnh32yeje7d3vkEzzcjevZfmT6F6omAvE
  priority: 102
  providerName: Wiley-Blackwell
Title Device Variation Effects on Neural Network Inference Accuracy in Analog In‐Memory Computing Systems
URI https://onlinelibrary.wiley.com/doi/abs/10.1002%2Faisy.202100199
https://www.proquest.com/docview/2703630292
https://doaj.org/article/350167ba1f1d481ab3ce6c5ff2adea55
Volume 4
WOSCitedRecordID wos000743074700001&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
journalDatabaseRights – providerCode: PRVAON
  databaseName: DOAJ: Directory of Open Access Journal (DOAJ)
  customDbUrl:
  eissn: 2640-4567
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0002171027
  issn: 2640-4567
  databaseCode: DOA
  dateStart: 20190101
  isFulltext: true
  titleUrlDefault: https://www.doaj.org/
  providerName: Directory of Open Access Journals
– providerCode: PRVHPJ
  databaseName: ROAD: Directory of Open Access Scholarly Resources (ISSN International Center)
  customDbUrl:
  eissn: 2640-4567
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0002171027
  issn: 2640-4567
  databaseCode: M~E
  dateStart: 20190101
  isFulltext: true
  titleUrlDefault: https://road.issn.org
  providerName: ISSN International Centre
– providerCode: PRVPQU
  databaseName: Advanced Technologies & Aerospace Database
  customDbUrl:
  eissn: 2640-4567
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0002171027
  issn: 2640-4567
  databaseCode: P5Z
  dateStart: 20201001
  isFulltext: true
  titleUrlDefault: https://search.proquest.com/hightechjournals
  providerName: ProQuest
– providerCode: PRVPQU
  databaseName: ProQuest Central
  customDbUrl:
  eissn: 2640-4567
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0002171027
  issn: 2640-4567
  databaseCode: BENPR
  dateStart: 20201001
  isFulltext: true
  titleUrlDefault: https://www.proquest.com/central
  providerName: ProQuest
– providerCode: PRVPQU
  databaseName: Publicly Available Content Database
  customDbUrl:
  eissn: 2640-4567
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0002171027
  issn: 2640-4567
  databaseCode: PIMPY
  dateStart: 20201001
  isFulltext: true
  titleUrlDefault: http://search.proquest.com/publiccontent
  providerName: ProQuest
– providerCode: PRVWIB
  databaseName: Wiley Online Library Free Content
  customDbUrl:
  eissn: 2640-4567
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0002171027
  issn: 2640-4567
  databaseCode: WIN
  dateStart: 20190101
  isFulltext: true
  titleUrlDefault: https://onlinelibrary.wiley.com
  providerName: Wiley-Blackwell
– providerCode: PRVWIB
  databaseName: Wiley Open Access
  customDbUrl:
  eissn: 2640-4567
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0002171027
  issn: 2640-4567
  databaseCode: 24P
  dateStart: 20190101
  isFulltext: true
  titleUrlDefault: https://authorservices.wiley.com/open-science/open-access/browse-journals.html
  providerName: Wiley-Blackwell
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV1Lb9QwELZoy4ELDwFiaVn5gMTJamLHiXNCW2jFSjSKeJSWi-VM7GqlardsWqS98RP4jfwSxo6TwgE4cIsfiRzPeGY8Hn9DyPOmkIAbtpJxMJJlLrGssVIxy01rIONNEdIBnbwtqkqdnpZ1dLh1MaxykIlBULcr8D7yfR6QohJe8peXX5jPGuVPV2MKjS2y41ESfOqGWn4efSxobqP-LAasxoTvm0W3wU0h98hDAe71RhcFyP7f7MxfrdWgbo7u_e9A75O70dCks54zHpBbdvmQ2NfWSwZ6gjvkQBLaoxd3FB89Tge-UfWB4XQ-XAWkMwBsgQ1dLKkHMVmdY-OPb9-PfZTuhvaJIVAF0gh__oh8PDr88OoNi4kWGAgpSqZEJtsiswoS53KTgCsSi9PV-gl0ZWFANGDbNCmNQnEo0tzhXwIaEzYz3IJ4TLaXq6V9QmiL8kAJ_EY4Kba84SY3Ao0cpyATyk0IG6ZcQ0Qh98kwLnSPn8y1J5EeSTQhL8b-lz3-xh97HngKjr08bnaoWK3PdVyG2h-j5kVjUpe2mUpNI8DmIJ1D1rRGygnZG4iq42Lu9A1FJ4QHnvjHUPRs_v5sLD39-zd3yR3ur1SEoMI9sn21vrbPyG34erXo1lOyxbN6SnYODqv63TR4DKaBybGunh_XZ1j6NK9-AvgcBSM
linkProvider ProQuest
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMw1V3LbtQwFL0qUyTY8BAgphTwAsTKamLnuUBooFSNOjMaiVK1K-Pc2NVIaKZMWtDs-AS-hI_iS7h2HoUFsOqCXRI7zsPH92FfnwvwrExjJIct5wJ1zCMbGF6aOONG6EpjJMrUpwM6GqfTaXZ8nM824Hu3F8aFVXYy0QvqaolujnxHeKaoQOTi1dkn7rJGudXVLoVGA4sDs_5CLlv9stil_n0uxN7bwzf7vM0qwFHGMueZjOIqjUyGgbWJDtCmgSG5UJFxHto81ShLNFUY5DqjsS_DxNKzkTSnibQwKKnda7AZObAPYHNWTGYn_ayOa4McvY4dMhA7el6vyQ0VjuvIE8xeaj-fJOA3y_ZX-9gruL3b_9uvuQO3WlOajRrs34UNs7gHZtc42ceO9KoBHWv4mWtGh46JhO6YNqHvrOg2O7IRIpXgms0XzNG0LE-p8MfXbxMXh7xmTeoLUvKsJXi_D--v5MsewGCxXJiHwCqSeJmkNvxauBGl0ImWZMbZDCOZ2SHwrosVtjzrLt3HR9UwRAvlIKF6SAzhRV__rGEY-WPN1w4xfS3HDO4vLFenqhU0yi0UJ2mpQxtWURbqUqJJMLaWBp_RcTyE7Q5EqhVXtbpE0BCEx-A_XkWNincn_dnW39t8Cjf2DydjNS6mB4_gpnAbSHwI5TYMzlcX5jFcx8_n83r1pB1ODD5cNUx_AloIXd4
linkToPdf http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMw1V3NbtQwEB6VghAXfgSIpQV8AHGKNrHjxDkgtLCsWLWsVgKqwsU4E7taCe22mwLaG4_A8_A4PAlj56dwAE49cEtix3Liz_Njj78BeFjmEslhKyKORkapi21UWqkiy01lMOVlHtIBHezns5k6PCzmW_C9Owvjwyo7mRgEdbVCv0Y-5IEpKuYFH7o2LGI-njw9Pol8Bim_09ql02ggsmc3X8h9q59MxzTWjzifvHjz_GXUZhiIUEhRREqksspTqzB2LjMxujy2JCMqMtQTV-QGRYm2SuLCKJIDIskc9QNJi9rUcIuC2r0AF3PyMX044Vy-79d3fAvk8nU8kTEfmkW9IYeUe9ajQDV7pgdDuoDfbNxfLeWg6ibX_uefdB2utgY2GzUz4gZs2eVNsGPrJSI7MOsGiqxhba4ZXXp-Enpj1gTEs2l3BJKNEKkEN2yxZJ68ZXVEhT--fnvlo5M3rEmIQaqftbTvt-DtuXzZbdherpb2DrCK5KAS1EbYIbe85CYzgow7pzAVyg0g6oZbY8u-7pOAfNQNbzTXHh66h8cAHvf1jxvekT_WfObR09fyfOHhwWp9pFvxo_32cZaXJnFJlarElAJthtI5mpLWSDmA3Q5QuhVitT5D0wB4wOM_uqJH09fv-ru7f2_zAVwmbOr96WxvB65wf6okxFXuwvbp-pO9B5fw8-miXt8P84rBh_PG6E-bcmVB
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%3Ajournal&rft.genre=article&rft.atitle=Device+Variation+Effects+on+Neural+Network+Inference+Accuracy+in+Analog+In%E2%80%90Memory+Computing+Systems&rft.jtitle=Advanced+intelligent+systems&rft.au=Wang%2C+Qiwen&rft.au=Park%2C+Yongmo&rft.au=Lu%2C+Wei+D.&rft.date=2022-08-01&rft.issn=2640-4567&rft.eissn=2640-4567&rft.volume=4&rft.issue=8&rft_id=info:doi/10.1002%2Faisy.202100199&rft.externalDBID=n%2Fa&rft.externalDocID=10_1002_aisy_202100199
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=2640-4567&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=2640-4567&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=2640-4567&client=summon