Accelerator for neural network computing and execution method thereof
An accelerator for neural network computing includes hardware engines and a buffer memory. The hardware engines include a convolution engine and at least a second engine. Each hardware engine includes circuitry to perform neural network operations. The buffer memory stores a first input tile and a s...
Gespeichert in:
| Hauptverfasser: | , , , , , , , , |
|---|---|
| Format: | Patent |
| Sprache: | chi ; eng |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | |
|---|---|
| container_issue | |
| container_start_page | |
| container_title | |
| container_volume | |
| creator | ANDRIAN, HENRRY LIN, CHIEN-HUNG HUNG, SHENG-JE WANG, SHAO-YU CHEN, TAI-LUNG CHEN, YI-SIOU WU, CHI-TA KUO, YU-TING CHENG, MENG-HSUAN |
| description | An accelerator for neural network computing includes hardware engines and a buffer memory. The hardware engines include a convolution engine and at least a second engine. Each hardware engine includes circuitry to perform neural network operations. The buffer memory stores a first input tile and a second input tile of an input feature map. The second input tile overlaps with the first input tile in the buffer memory. The convolution engine is operative to retrieve the first input tile from the buffer memory, perform convolution operations on the first input tile to generate an intermediate tile of an intermediate feature map, and pass the intermediate tile to the second engine via the buffer memory. |
| format | Patent |
| fullrecord | <record><control><sourceid>epo_EVB</sourceid><recordid>TN_cdi_epo_espacenet_TWI748151BB</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>TWI748151BB</sourcerecordid><originalsourceid>FETCH-epo_espacenet_TWI748151BB3</originalsourceid><addsrcrecordid>eNrjZHB1TE5OzUktSizJL1JIA-K81NKixBwgVVKeX5StkJyfW1BakpmXrpCYl6KQWpGaDOTl5ynkppZk5KcolGSkFqXmp_EwsKYl5hSn8kJpbgYFN9cQZw_d1IL8-NTigsTkVKCB8SHhnuYmFoamhk5OxkQoAQDY_TPS</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>patent</recordtype></control><display><type>patent</type><title>Accelerator for neural network computing and execution method thereof</title><source>esp@cenet</source><creator>ANDRIAN, HENRRY ; LIN, CHIEN-HUNG ; HUNG, SHENG-JE ; WANG, SHAO-YU ; CHEN, TAI-LUNG ; CHEN, YI-SIOU ; WU, CHI-TA ; KUO, YU-TING ; CHENG, MENG-HSUAN</creator><creatorcontrib>ANDRIAN, HENRRY ; LIN, CHIEN-HUNG ; HUNG, SHENG-JE ; WANG, SHAO-YU ; CHEN, TAI-LUNG ; CHEN, YI-SIOU ; WU, CHI-TA ; KUO, YU-TING ; CHENG, MENG-HSUAN</creatorcontrib><description>An accelerator for neural network computing includes hardware engines and a buffer memory. The hardware engines include a convolution engine and at least a second engine. Each hardware engine includes circuitry to perform neural network operations. The buffer memory stores a first input tile and a second input tile of an input feature map. The second input tile overlaps with the first input tile in the buffer memory. The convolution engine is operative to retrieve the first input tile from the buffer memory, perform convolution operations on the first input tile to generate an intermediate tile of an intermediate feature map, and pass the intermediate tile to the second engine via the buffer memory.</description><language>chi ; eng</language><subject>CALCULATING ; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS ; COMPUTING ; COUNTING ; ELECTRIC DIGITAL DATA PROCESSING ; PHYSICS</subject><creationdate>2021</creationdate><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://worldwide.espacenet.com/publicationDetails/biblio?FT=D&date=20211201&DB=EPODOC&CC=TW&NR=I748151B$$EHTML$$P50$$Gepo$$Hfree_for_read</linktohtml><link.rule.ids>230,308,776,881,25543,76292</link.rule.ids><linktorsrc>$$Uhttps://worldwide.espacenet.com/publicationDetails/biblio?FT=D&date=20211201&DB=EPODOC&CC=TW&NR=I748151B$$EView_record_in_European_Patent_Office$$FView_record_in_$$GEuropean_Patent_Office$$Hfree_for_read</linktorsrc></links><search><creatorcontrib>ANDRIAN, HENRRY</creatorcontrib><creatorcontrib>LIN, CHIEN-HUNG</creatorcontrib><creatorcontrib>HUNG, SHENG-JE</creatorcontrib><creatorcontrib>WANG, SHAO-YU</creatorcontrib><creatorcontrib>CHEN, TAI-LUNG</creatorcontrib><creatorcontrib>CHEN, YI-SIOU</creatorcontrib><creatorcontrib>WU, CHI-TA</creatorcontrib><creatorcontrib>KUO, YU-TING</creatorcontrib><creatorcontrib>CHENG, MENG-HSUAN</creatorcontrib><title>Accelerator for neural network computing and execution method thereof</title><description>An accelerator for neural network computing includes hardware engines and a buffer memory. The hardware engines include a convolution engine and at least a second engine. Each hardware engine includes circuitry to perform neural network operations. The buffer memory stores a first input tile and a second input tile of an input feature map. The second input tile overlaps with the first input tile in the buffer memory. The convolution engine is operative to retrieve the first input tile from the buffer memory, perform convolution operations on the first input tile to generate an intermediate tile of an intermediate feature map, and pass the intermediate tile to the second engine via the buffer memory.</description><subject>CALCULATING</subject><subject>COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS</subject><subject>COMPUTING</subject><subject>COUNTING</subject><subject>ELECTRIC DIGITAL DATA PROCESSING</subject><subject>PHYSICS</subject><fulltext>true</fulltext><rsrctype>patent</rsrctype><creationdate>2021</creationdate><recordtype>patent</recordtype><sourceid>EVB</sourceid><recordid>eNrjZHB1TE5OzUktSizJL1JIA-K81NKixBwgVVKeX5StkJyfW1BakpmXrpCYl6KQWpGaDOTl5ynkppZk5KcolGSkFqXmp_EwsKYl5hSn8kJpbgYFN9cQZw_d1IL8-NTigsTkVKCB8SHhnuYmFoamhk5OxkQoAQDY_TPS</recordid><startdate>20211201</startdate><enddate>20211201</enddate><creator>ANDRIAN, HENRRY</creator><creator>LIN, CHIEN-HUNG</creator><creator>HUNG, SHENG-JE</creator><creator>WANG, SHAO-YU</creator><creator>CHEN, TAI-LUNG</creator><creator>CHEN, YI-SIOU</creator><creator>WU, CHI-TA</creator><creator>KUO, YU-TING</creator><creator>CHENG, MENG-HSUAN</creator><scope>EVB</scope></search><sort><creationdate>20211201</creationdate><title>Accelerator for neural network computing and execution method thereof</title><author>ANDRIAN, HENRRY ; LIN, CHIEN-HUNG ; HUNG, SHENG-JE ; WANG, SHAO-YU ; CHEN, TAI-LUNG ; CHEN, YI-SIOU ; WU, CHI-TA ; KUO, YU-TING ; CHENG, MENG-HSUAN</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-epo_espacenet_TWI748151BB3</frbrgroupid><rsrctype>patents</rsrctype><prefilter>patents</prefilter><language>chi ; eng</language><creationdate>2021</creationdate><topic>CALCULATING</topic><topic>COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS</topic><topic>COMPUTING</topic><topic>COUNTING</topic><topic>ELECTRIC DIGITAL DATA PROCESSING</topic><topic>PHYSICS</topic><toplevel>online_resources</toplevel><creatorcontrib>ANDRIAN, HENRRY</creatorcontrib><creatorcontrib>LIN, CHIEN-HUNG</creatorcontrib><creatorcontrib>HUNG, SHENG-JE</creatorcontrib><creatorcontrib>WANG, SHAO-YU</creatorcontrib><creatorcontrib>CHEN, TAI-LUNG</creatorcontrib><creatorcontrib>CHEN, YI-SIOU</creatorcontrib><creatorcontrib>WU, CHI-TA</creatorcontrib><creatorcontrib>KUO, YU-TING</creatorcontrib><creatorcontrib>CHENG, MENG-HSUAN</creatorcontrib><collection>esp@cenet</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>ANDRIAN, HENRRY</au><au>LIN, CHIEN-HUNG</au><au>HUNG, SHENG-JE</au><au>WANG, SHAO-YU</au><au>CHEN, TAI-LUNG</au><au>CHEN, YI-SIOU</au><au>WU, CHI-TA</au><au>KUO, YU-TING</au><au>CHENG, MENG-HSUAN</au><format>patent</format><genre>patent</genre><ristype>GEN</ristype><title>Accelerator for neural network computing and execution method thereof</title><date>2021-12-01</date><risdate>2021</risdate><abstract>An accelerator for neural network computing includes hardware engines and a buffer memory. The hardware engines include a convolution engine and at least a second engine. Each hardware engine includes circuitry to perform neural network operations. The buffer memory stores a first input tile and a second input tile of an input feature map. The second input tile overlaps with the first input tile in the buffer memory. The convolution engine is operative to retrieve the first input tile from the buffer memory, perform convolution operations on the first input tile to generate an intermediate tile of an intermediate feature map, and pass the intermediate tile to the second engine via the buffer memory.</abstract><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | |
| ispartof | |
| issn | |
| language | chi ; eng |
| recordid | cdi_epo_espacenet_TWI748151BB |
| source | esp@cenet |
| subjects | CALCULATING COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS COMPUTING COUNTING ELECTRIC DIGITAL DATA PROCESSING PHYSICS |
| title | Accelerator for neural network computing and execution method thereof |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-27T09%3A06%3A32IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-epo_EVB&rft_val_fmt=info:ofi/fmt:kev:mtx:patent&rft.genre=patent&rft.au=ANDRIAN,%20HENRRY&rft.date=2021-12-01&rft_id=info:doi/&rft_dat=%3Cepo_EVB%3ETWI748151BB%3C/epo_EVB%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true |