A More Hardware-Oriented Spiking Neural Network Based on Leading Memory Technology and Its Application With Reinforcement Learning
In recent days, more hardware-driven artificial intelligence system capable of brain-like low-energy consumption is gaining ever-increasing interest. The hardware-driven property lies in the low-power synaptic device and its array along with the area and energy-efficient neuron circuits. In this wor...
Gespeichert in:
| Veröffentlicht in: | IEEE transactions on electron devices 2021-09, Vol.68 (9), p.4411-4417 |
|---|---|
| Hauptverfasser: | , , , , , , , |
| Format: | Artikel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | 4417 |
|---|---|
| container_issue | 9 |
| container_start_page | 4411 |
| container_title | IEEE transactions on electron devices |
| container_volume | 68 |
| creator | Kim, Min-Hwi Hwang, Sungmin Bang, Suhyun Kim, Tae-Hyeon Lee, Dong Keun Ansari, Md. Hasan Raza Cho, Seongjae Park, Byung-Gook |
| description | In recent days, more hardware-driven artificial intelligence system capable of brain-like low-energy consumption is gaining ever-increasing interest. The hardware-driven property lies in the low-power synaptic device and its array along with the area and energy-efficient neuron circuits. In this work, a spiking neural network (SNN) based on analog synaptic device of resistive-switching random access memory (RRAM) is constructed from the experimentally fabricated devices. Furthermore, the capability of the designed SNN hardware for sequential tasks through an optimal reinforcement learning (RL) algorithm is demonstrated. More specifically, the Rush Hour game is conducted as an example of applications for the sequential task for which an SNN architecture is plausibly suited. The rule of the game is simple but has not been demonstrated by a hardware-oriented artificial neural network (ANN) yet, and in this work, it is reported that the analog RRAM synaptic devices in the cross-point array architecture successfully solve the problem via the RL algorithm. |
| doi_str_mv | 10.1109/TED.2021.3099769 |
| format | Article |
| fullrecord | <record><control><sourceid>proquest_RIE</sourceid><recordid>TN_cdi_crossref_primary_10_1109_TED_2021_3099769</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>9506994</ieee_id><sourcerecordid>2562950273</sourcerecordid><originalsourceid>FETCH-LOGICAL-c291t-ec09576f84215c2a70252faf4b7507c70c8132369f0caab7f65e45c7b42242dc3</originalsourceid><addsrcrecordid>eNo9kM1PAjEUxBujiYjeTbw08bzYdvtBj4goJCCJYjxuSvctlI_t2l1CuPqXWwLxNHl5v5lJBqF7SjqUEv00G7x0GGG0kxKtldQXqEWFUImWXF6iFiG0m-i0m16jm7pexVNyzlrot4cnPgAempDvTYBkGhyUDeT4s3JrVy7wO-yC2URp9j6s8bOp49OXeAwmP_4nsPXhgGdgl6Xf-MUBmzLHo6bGvaraOGsaF-lv1yzxB7iy8MHCNlYcA0IZE27RVWE2NdydtY2-Xgez_jAZT99G_d44sUzTJgFLtFCy6HJGhWVGESZYYQo-V4Ioq4jt0pSlUhfEGjNXhRTAhVVzzhhnuU3b6PGUWwX_s4O6yVZ-F8pYmTEhmRaEqTRS5ETZ4Os6QJFVwW1NOGSUZMels7h0dlw6Oy8dLQ8niwOAfzzmSa15-gdQ-3pa</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2562950273</pqid></control><display><type>article</type><title>A More Hardware-Oriented Spiking Neural Network Based on Leading Memory Technology and Its Application With Reinforcement Learning</title><source>IEEE Electronic Library (IEL)</source><creator>Kim, Min-Hwi ; Hwang, Sungmin ; Bang, Suhyun ; Kim, Tae-Hyeon ; Lee, Dong Keun ; Ansari, Md. Hasan Raza ; Cho, Seongjae ; Park, Byung-Gook</creator><creatorcontrib>Kim, Min-Hwi ; Hwang, Sungmin ; Bang, Suhyun ; Kim, Tae-Hyeon ; Lee, Dong Keun ; Ansari, Md. Hasan Raza ; Cho, Seongjae ; Park, Byung-Gook</creatorcontrib><description>In recent days, more hardware-driven artificial intelligence system capable of brain-like low-energy consumption is gaining ever-increasing interest. The hardware-driven property lies in the low-power synaptic device and its array along with the area and energy-efficient neuron circuits. In this work, a spiking neural network (SNN) based on analog synaptic device of resistive-switching random access memory (RRAM) is constructed from the experimentally fabricated devices. Furthermore, the capability of the designed SNN hardware for sequential tasks through an optimal reinforcement learning (RL) algorithm is demonstrated. More specifically, the Rush Hour game is conducted as an example of applications for the sequential task for which an SNN architecture is plausibly suited. The rule of the game is simple but has not been demonstrated by a hardware-oriented artificial neural network (ANN) yet, and in this work, it is reported that the analog RRAM synaptic devices in the cross-point array architecture successfully solve the problem via the RL algorithm.</description><identifier>ISSN: 0018-9383</identifier><identifier>EISSN: 1557-9646</identifier><identifier>DOI: 10.1109/TED.2021.3099769</identifier><identifier>CODEN: IETDAI</identifier><language>eng</language><publisher>New York: IEEE</publisher><subject>Algorithms ; Arrays ; Artificial intelligence ; Artificial neural network (ANN) ; Artificial neural networks ; Biological neural networks ; Computer architecture ; cross-point array architecture ; Energy consumption ; Games ; Hardware ; hardware-driven artificial intelligence ; Learning theory ; low energy consumption ; Machine learning ; Neural networks ; Neurons ; Random access memory ; reinforcement learning (RL) ; resistive-switching random access memory (RRAM) ; Rush Hour game ; sequential task ; Silicon ; Silicon compounds ; Spiking ; spiking neural network (SNN) ; Switches ; synaptic device</subject><ispartof>IEEE transactions on electron devices, 2021-09, Vol.68 (9), p.4411-4417</ispartof><rights>Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2021</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c291t-ec09576f84215c2a70252faf4b7507c70c8132369f0caab7f65e45c7b42242dc3</citedby><cites>FETCH-LOGICAL-c291t-ec09576f84215c2a70252faf4b7507c70c8132369f0caab7f65e45c7b42242dc3</cites><orcidid>0000-0002-2962-2458 ; 0000-0003-0964-7583 ; 0000-0001-8520-718X ; 0000-0002-2617-7627 ; 0000-0002-8587-4588 ; 0000-0001-8347-8656 ; 0000-0002-7519-4896</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/9506994$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>314,777,781,793,27905,27906,54739</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/9506994$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Kim, Min-Hwi</creatorcontrib><creatorcontrib>Hwang, Sungmin</creatorcontrib><creatorcontrib>Bang, Suhyun</creatorcontrib><creatorcontrib>Kim, Tae-Hyeon</creatorcontrib><creatorcontrib>Lee, Dong Keun</creatorcontrib><creatorcontrib>Ansari, Md. Hasan Raza</creatorcontrib><creatorcontrib>Cho, Seongjae</creatorcontrib><creatorcontrib>Park, Byung-Gook</creatorcontrib><title>A More Hardware-Oriented Spiking Neural Network Based on Leading Memory Technology and Its Application With Reinforcement Learning</title><title>IEEE transactions on electron devices</title><addtitle>TED</addtitle><description>In recent days, more hardware-driven artificial intelligence system capable of brain-like low-energy consumption is gaining ever-increasing interest. The hardware-driven property lies in the low-power synaptic device and its array along with the area and energy-efficient neuron circuits. In this work, a spiking neural network (SNN) based on analog synaptic device of resistive-switching random access memory (RRAM) is constructed from the experimentally fabricated devices. Furthermore, the capability of the designed SNN hardware for sequential tasks through an optimal reinforcement learning (RL) algorithm is demonstrated. More specifically, the Rush Hour game is conducted as an example of applications for the sequential task for which an SNN architecture is plausibly suited. The rule of the game is simple but has not been demonstrated by a hardware-oriented artificial neural network (ANN) yet, and in this work, it is reported that the analog RRAM synaptic devices in the cross-point array architecture successfully solve the problem via the RL algorithm.</description><subject>Algorithms</subject><subject>Arrays</subject><subject>Artificial intelligence</subject><subject>Artificial neural network (ANN)</subject><subject>Artificial neural networks</subject><subject>Biological neural networks</subject><subject>Computer architecture</subject><subject>cross-point array architecture</subject><subject>Energy consumption</subject><subject>Games</subject><subject>Hardware</subject><subject>hardware-driven artificial intelligence</subject><subject>Learning theory</subject><subject>low energy consumption</subject><subject>Machine learning</subject><subject>Neural networks</subject><subject>Neurons</subject><subject>Random access memory</subject><subject>reinforcement learning (RL)</subject><subject>resistive-switching random access memory (RRAM)</subject><subject>Rush Hour game</subject><subject>sequential task</subject><subject>Silicon</subject><subject>Silicon compounds</subject><subject>Spiking</subject><subject>spiking neural network (SNN)</subject><subject>Switches</subject><subject>synaptic device</subject><issn>0018-9383</issn><issn>1557-9646</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2021</creationdate><recordtype>article</recordtype><sourceid>RIE</sourceid><recordid>eNo9kM1PAjEUxBujiYjeTbw08bzYdvtBj4goJCCJYjxuSvctlI_t2l1CuPqXWwLxNHl5v5lJBqF7SjqUEv00G7x0GGG0kxKtldQXqEWFUImWXF6iFiG0m-i0m16jm7pexVNyzlrot4cnPgAempDvTYBkGhyUDeT4s3JrVy7wO-yC2URp9j6s8bOp49OXeAwmP_4nsPXhgGdgl6Xf-MUBmzLHo6bGvaraOGsaF-lv1yzxB7iy8MHCNlYcA0IZE27RVWE2NdydtY2-Xgez_jAZT99G_d44sUzTJgFLtFCy6HJGhWVGESZYYQo-V4Ioq4jt0pSlUhfEGjNXhRTAhVVzzhhnuU3b6PGUWwX_s4O6yVZ-F8pYmTEhmRaEqTRS5ETZ4Os6QJFVwW1NOGSUZMels7h0dlw6Oy8dLQ8niwOAfzzmSa15-gdQ-3pa</recordid><startdate>20210901</startdate><enddate>20210901</enddate><creator>Kim, Min-Hwi</creator><creator>Hwang, Sungmin</creator><creator>Bang, Suhyun</creator><creator>Kim, Tae-Hyeon</creator><creator>Lee, Dong Keun</creator><creator>Ansari, Md. Hasan Raza</creator><creator>Cho, Seongjae</creator><creator>Park, Byung-Gook</creator><general>IEEE</general><general>The Institute of Electrical and Electronics Engineers, Inc. (IEEE)</general><scope>97E</scope><scope>RIA</scope><scope>RIE</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7SP</scope><scope>8FD</scope><scope>L7M</scope><orcidid>https://orcid.org/0000-0002-2962-2458</orcidid><orcidid>https://orcid.org/0000-0003-0964-7583</orcidid><orcidid>https://orcid.org/0000-0001-8520-718X</orcidid><orcidid>https://orcid.org/0000-0002-2617-7627</orcidid><orcidid>https://orcid.org/0000-0002-8587-4588</orcidid><orcidid>https://orcid.org/0000-0001-8347-8656</orcidid><orcidid>https://orcid.org/0000-0002-7519-4896</orcidid></search><sort><creationdate>20210901</creationdate><title>A More Hardware-Oriented Spiking Neural Network Based on Leading Memory Technology and Its Application With Reinforcement Learning</title><author>Kim, Min-Hwi ; Hwang, Sungmin ; Bang, Suhyun ; Kim, Tae-Hyeon ; Lee, Dong Keun ; Ansari, Md. Hasan Raza ; Cho, Seongjae ; Park, Byung-Gook</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c291t-ec09576f84215c2a70252faf4b7507c70c8132369f0caab7f65e45c7b42242dc3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2021</creationdate><topic>Algorithms</topic><topic>Arrays</topic><topic>Artificial intelligence</topic><topic>Artificial neural network (ANN)</topic><topic>Artificial neural networks</topic><topic>Biological neural networks</topic><topic>Computer architecture</topic><topic>cross-point array architecture</topic><topic>Energy consumption</topic><topic>Games</topic><topic>Hardware</topic><topic>hardware-driven artificial intelligence</topic><topic>Learning theory</topic><topic>low energy consumption</topic><topic>Machine learning</topic><topic>Neural networks</topic><topic>Neurons</topic><topic>Random access memory</topic><topic>reinforcement learning (RL)</topic><topic>resistive-switching random access memory (RRAM)</topic><topic>Rush Hour game</topic><topic>sequential task</topic><topic>Silicon</topic><topic>Silicon compounds</topic><topic>Spiking</topic><topic>spiking neural network (SNN)</topic><topic>Switches</topic><topic>synaptic device</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Kim, Min-Hwi</creatorcontrib><creatorcontrib>Hwang, Sungmin</creatorcontrib><creatorcontrib>Bang, Suhyun</creatorcontrib><creatorcontrib>Kim, Tae-Hyeon</creatorcontrib><creatorcontrib>Lee, Dong Keun</creatorcontrib><creatorcontrib>Ansari, Md. Hasan Raza</creatorcontrib><creatorcontrib>Cho, Seongjae</creatorcontrib><creatorcontrib>Park, Byung-Gook</creatorcontrib><collection>IEEE All-Society Periodicals Package (ASPP) 2005-present</collection><collection>IEEE All-Society Periodicals Package (ASPP) 1998-Present</collection><collection>IEEE Electronic Library (IEL)</collection><collection>CrossRef</collection><collection>Electronics & Communications Abstracts</collection><collection>Technology Research Database</collection><collection>Advanced Technologies Database with Aerospace</collection><jtitle>IEEE transactions on electron devices</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Kim, Min-Hwi</au><au>Hwang, Sungmin</au><au>Bang, Suhyun</au><au>Kim, Tae-Hyeon</au><au>Lee, Dong Keun</au><au>Ansari, Md. Hasan Raza</au><au>Cho, Seongjae</au><au>Park, Byung-Gook</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>A More Hardware-Oriented Spiking Neural Network Based on Leading Memory Technology and Its Application With Reinforcement Learning</atitle><jtitle>IEEE transactions on electron devices</jtitle><stitle>TED</stitle><date>2021-09-01</date><risdate>2021</risdate><volume>68</volume><issue>9</issue><spage>4411</spage><epage>4417</epage><pages>4411-4417</pages><issn>0018-9383</issn><eissn>1557-9646</eissn><coden>IETDAI</coden><abstract>In recent days, more hardware-driven artificial intelligence system capable of brain-like low-energy consumption is gaining ever-increasing interest. The hardware-driven property lies in the low-power synaptic device and its array along with the area and energy-efficient neuron circuits. In this work, a spiking neural network (SNN) based on analog synaptic device of resistive-switching random access memory (RRAM) is constructed from the experimentally fabricated devices. Furthermore, the capability of the designed SNN hardware for sequential tasks through an optimal reinforcement learning (RL) algorithm is demonstrated. More specifically, the Rush Hour game is conducted as an example of applications for the sequential task for which an SNN architecture is plausibly suited. The rule of the game is simple but has not been demonstrated by a hardware-oriented artificial neural network (ANN) yet, and in this work, it is reported that the analog RRAM synaptic devices in the cross-point array architecture successfully solve the problem via the RL algorithm.</abstract><cop>New York</cop><pub>IEEE</pub><doi>10.1109/TED.2021.3099769</doi><tpages>7</tpages><orcidid>https://orcid.org/0000-0002-2962-2458</orcidid><orcidid>https://orcid.org/0000-0003-0964-7583</orcidid><orcidid>https://orcid.org/0000-0001-8520-718X</orcidid><orcidid>https://orcid.org/0000-0002-2617-7627</orcidid><orcidid>https://orcid.org/0000-0002-8587-4588</orcidid><orcidid>https://orcid.org/0000-0001-8347-8656</orcidid><orcidid>https://orcid.org/0000-0002-7519-4896</orcidid></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | ISSN: 0018-9383 |
| ispartof | IEEE transactions on electron devices, 2021-09, Vol.68 (9), p.4411-4417 |
| issn | 0018-9383 1557-9646 |
| language | eng |
| recordid | cdi_crossref_primary_10_1109_TED_2021_3099769 |
| source | IEEE Electronic Library (IEL) |
| subjects | Algorithms Arrays Artificial intelligence Artificial neural network (ANN) Artificial neural networks Biological neural networks Computer architecture cross-point array architecture Energy consumption Games Hardware hardware-driven artificial intelligence Learning theory low energy consumption Machine learning Neural networks Neurons Random access memory reinforcement learning (RL) resistive-switching random access memory (RRAM) Rush Hour game sequential task Silicon Silicon compounds Spiking spiking neural network (SNN) Switches synaptic device |
| title | A More Hardware-Oriented Spiking Neural Network Based on Leading Memory Technology and Its Application With Reinforcement Learning |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-20T10%3A00%3A29IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_RIE&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=A%20More%20Hardware-Oriented%20Spiking%20Neural%20Network%20Based%20on%20Leading%20Memory%20Technology%20and%20Its%20Application%20With%20Reinforcement%20Learning&rft.jtitle=IEEE%20transactions%20on%20electron%20devices&rft.au=Kim,%20Min-Hwi&rft.date=2021-09-01&rft.volume=68&rft.issue=9&rft.spage=4411&rft.epage=4417&rft.pages=4411-4417&rft.issn=0018-9383&rft.eissn=1557-9646&rft.coden=IETDAI&rft_id=info:doi/10.1109/TED.2021.3099769&rft_dat=%3Cproquest_RIE%3E2562950273%3C/proquest_RIE%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2562950273&rft_id=info:pmid/&rft_ieee_id=9506994&rfr_iscdi=true |