{"id":"https://openalex.org/W4392265952","doi":"https://doi.org/10.1109/cgo57630.2024.10444805","title":"TapeFlow: Streaming Gradient Tapes in Automatic Differentiation","display_name":"TapeFlow: Streaming Gradient Tapes in Automatic Differentiation","publication_year":2024,"publication_date":"2024-02-28","ids":{"openalex":"https://openalex.org/W4392265952","doi":"https://doi.org/10.1109/cgo57630.2024.10444805"},"language":"en","primary_location":{"id":"doi:10.1109/cgo57630.2024.10444805","is_oa":false,"landing_page_url":"http://dx.doi.org/10.1109/cgo57630.2024.10444805","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 IEEE/ACM International Symposium on Code Generation and Optimization (CGO)","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5000642432","display_name":"Milad Hakimi","orcid":null},"institutions":[{"id":"https://openalex.org/I18014758","display_name":"Simon Fraser University","ror":"https://ror.org/0213rcc28","country_code":"CA","type":"education","lineage":["https://openalex.org/I18014758"]}],"countries":["CA"],"is_corresponding":true,"raw_author_name":"Milad Hakimi","raw_affiliation_strings":["School of Computing Sciences, Simon Fraser University,Burnaby,Canada","School of Computing Sciences, Simon Fraser University, Burnaby, Canada"],"affiliations":[{"raw_affiliation_string":"School of Computing Sciences, Simon Fraser University,Burnaby,Canada","institution_ids":["https://openalex.org/I18014758"]},{"raw_affiliation_string":"School of Computing Sciences, Simon Fraser University, Burnaby, Canada","institution_ids":["https://openalex.org/I18014758"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5035046431","display_name":"Arrvindh Shriraman","orcid":"https://orcid.org/0000-0001-8466-5701"},"institutions":[{"id":"https://openalex.org/I18014758","display_name":"Simon Fraser University","ror":"https://ror.org/0213rcc28","country_code":"CA","type":"education","lineage":["https://openalex.org/I18014758"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Arrvindh Shriraman","raw_affiliation_strings":["School of Computing Sciences, Simon Fraser University,Burnaby,Canada","School of Computing Sciences, Simon Fraser University, Burnaby, Canada"],"affiliations":[{"raw_affiliation_string":"School of Computing Sciences, Simon Fraser University,Burnaby,Canada","institution_ids":["https://openalex.org/I18014758"]},{"raw_affiliation_string":"School of Computing Sciences, Simon Fraser University, Burnaby, Canada","institution_ids":["https://openalex.org/I18014758"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5000642432"],"corresponding_institution_ids":["https://openalex.org/I18014758"],"apc_list":null,"apc_paid":null,"fwci":1.0638,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.72176631,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":95},"biblio":{"volume":"57","issue":null,"first_page":"81","last_page":"92"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9997000098228455,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9997000098228455,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9915000200271606,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11181","display_name":"Advanced Data Storage Technologies","score":0.9896000027656555,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8486308455467224},{"id":"https://openalex.org/keywords/operand","display_name":"Operand","score":0.7825771570205688},{"id":"https://openalex.org/keywords/compiler","display_name":"Compiler","score":0.7189997434616089},{"id":"https://openalex.org/keywords/cache","display_name":"Cache","score":0.701142430305481},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.6763884425163269},{"id":"https://openalex.org/keywords/stencil","display_name":"Stencil","score":0.6404362320899963},{"id":"https://openalex.org/keywords/static-random-access-memory","display_name":"Static random-access memory","score":0.5294952988624573},{"id":"https://openalex.org/keywords/reuse","display_name":"Reuse","score":0.5207479000091553},{"id":"https://openalex.org/keywords/schedule","display_name":"Schedule","score":0.50240159034729},{"id":"https://openalex.org/keywords/locality-of-reference","display_name":"Locality of reference","score":0.4747781455516815},{"id":"https://openalex.org/keywords/dram","display_name":"Dram","score":0.4643877446651459},{"id":"https://openalex.org/keywords/code","display_name":"Code (set theory)","score":0.4428033232688904},{"id":"https://openalex.org/keywords/overlay","display_name":"Overlay","score":0.43139025568962097},{"id":"https://openalex.org/keywords/chip","display_name":"Chip","score":0.4269981384277344},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.41857272386550903},{"id":"https://openalex.org/keywords/optimizing-compiler","display_name":"Optimizing compiler","score":0.41818007826805115},{"id":"https://openalex.org/keywords/automatic-differentiation","display_name":"Automatic differentiation","score":0.4135648012161255},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.3984748125076294},{"id":"https://openalex.org/keywords/computer-hardware","display_name":"Computer hardware","score":0.3295402526855469},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.28534573316574097},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.2716038227081299},{"id":"https://openalex.org/keywords/computational-science","display_name":"Computational science","score":0.24696511030197144},{"id":"https://openalex.org/keywords/programming-language","display_name":"Programming language","score":0.2186695635318756},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.21268680691719055}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8486308455467224},{"id":"https://openalex.org/C55526617","wikidata":"https://www.wikidata.org/wiki/Q719375","display_name":"Operand","level":2,"score":0.7825771570205688},{"id":"https://openalex.org/C169590947","wikidata":"https://www.wikidata.org/wiki/Q47506","display_name":"Compiler","level":2,"score":0.7189997434616089},{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.701142430305481},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.6763884425163269},{"id":"https://openalex.org/C76752949","wikidata":"https://www.wikidata.org/wiki/Q7607499","display_name":"Stencil","level":2,"score":0.6404362320899963},{"id":"https://openalex.org/C68043766","wikidata":"https://www.wikidata.org/wiki/Q267416","display_name":"Static random-access memory","level":2,"score":0.5294952988624573},{"id":"https://openalex.org/C206588197","wikidata":"https://www.wikidata.org/wiki/Q846574","display_name":"Reuse","level":2,"score":0.5207479000091553},{"id":"https://openalex.org/C68387754","wikidata":"https://www.wikidata.org/wiki/Q7271585","display_name":"Schedule","level":2,"score":0.50240159034729},{"id":"https://openalex.org/C27602214","wikidata":"https://www.wikidata.org/wiki/Q1868547","display_name":"Locality of reference","level":3,"score":0.4747781455516815},{"id":"https://openalex.org/C7366592","wikidata":"https://www.wikidata.org/wiki/Q1255620","display_name":"Dram","level":2,"score":0.4643877446651459},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.4428033232688904},{"id":"https://openalex.org/C136085584","wikidata":"https://www.wikidata.org/wiki/Q910289","display_name":"Overlay","level":2,"score":0.43139025568962097},{"id":"https://openalex.org/C165005293","wikidata":"https://www.wikidata.org/wiki/Q1074500","display_name":"Chip","level":2,"score":0.4269981384277344},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.41857272386550903},{"id":"https://openalex.org/C190902152","wikidata":"https://www.wikidata.org/wiki/Q1325106","display_name":"Optimizing compiler","level":3,"score":0.41818007826805115},{"id":"https://openalex.org/C133512626","wikidata":"https://www.wikidata.org/wiki/Q787371","display_name":"Automatic differentiation","level":3,"score":0.4135648012161255},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.3984748125076294},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.3295402526855469},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.28534573316574097},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2716038227081299},{"id":"https://openalex.org/C459310","wikidata":"https://www.wikidata.org/wiki/Q117801","display_name":"Computational science","level":1,"score":0.24696511030197144},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.2186695635318756},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.21268680691719055},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0},{"id":"https://openalex.org/C18903297","wikidata":"https://www.wikidata.org/wiki/Q7150","display_name":"Ecology","level":1,"score":0.0},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/cgo57630.2024.10444805","is_oa":false,"landing_page_url":"http://dx.doi.org/10.1109/cgo57630.2024.10444805","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 IEEE/ACM International Symposium on Code Generation and Optimization (CGO)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":46,"referenced_works":["https://openalex.org/W400678491","https://openalex.org/W1498436455","https://openalex.org/W1965256357","https://openalex.org/W1966377066","https://openalex.org/W2031070741","https://openalex.org/W2032904810","https://openalex.org/W2049797245","https://openalex.org/W2105102111","https://openalex.org/W2105311743","https://openalex.org/W2239894058","https://openalex.org/W2387022868","https://openalex.org/W2556096037","https://openalex.org/W2590246587","https://openalex.org/W2626953429","https://openalex.org/W2747329762","https://openalex.org/W2810610794","https://openalex.org/W2883283076","https://openalex.org/W2883830791","https://openalex.org/W2899771611","https://openalex.org/W2944824859","https://openalex.org/W2966295917","https://openalex.org/W2969388332","https://openalex.org/W2972073717","https://openalex.org/W3012514909","https://openalex.org/W3101104221","https://openalex.org/W3102724434","https://openalex.org/W3106090013","https://openalex.org/W3106346928","https://openalex.org/W3123635906","https://openalex.org/W3129831491","https://openalex.org/W3157864729","https://openalex.org/W3179431530","https://openalex.org/W3210510236","https://openalex.org/W4240636081","https://openalex.org/W4286961957","https://openalex.org/W4293404878","https://openalex.org/W4295312788","https://openalex.org/W4298399544","https://openalex.org/W4301361180","https://openalex.org/W6713134421","https://openalex.org/W6766978945","https://openalex.org/W6769017845","https://openalex.org/W6775855194","https://openalex.org/W6784424872","https://openalex.org/W6786062066","https://openalex.org/W6801585611"],"related_works":["https://openalex.org/W3102564634","https://openalex.org/W2129829929","https://openalex.org/W4384755779","https://openalex.org/W2552708242","https://openalex.org/W2010557350","https://openalex.org/W114331401","https://openalex.org/W2104840748","https://openalex.org/W2123433857","https://openalex.org/W2000618606","https://openalex.org/W2167013139"],"abstract_inverted_index":{"Computing":[0],"gradients":[1,21],"is":[2,35,47],"a":[3,60,76,125,149,190],"crucial":[4],"task":[5],"in":[6,53,196],"many":[7],"domains,":[8],"including":[9],"machine":[10],"learning,":[11],"physics":[12],"simulations,":[13],"and":[14,67,81,118,133,181],"scientific":[15],"computing.":[16],"Automatic":[17],"differentiation":[18],"(AD)":[19],"computes":[20],"for":[22,42,158],"arbitrary":[23],"imperative":[24],"code.":[25],"In":[26],"reverse":[27],"mode":[28],"AD,":[29],"an":[30,104],"auxiliary":[31],"structure,":[32],"the":[33,51,54,83,94,100,114,121,129,140,143,164,169],"tape,":[34],"used":[36],"to":[37,49,79,107],"transfer":[38],"intermediary":[39],"values":[40],"required":[41,130],"gradient":[43,84],"computation.":[44],"The":[45,153],"challenge":[46],"how":[48],"organize":[50],"tape":[52,101,122,141,154],"memory":[55],"hierarchy":[56],"since":[57],"it":[58,147],"has":[59],"high":[61],"reuse":[62],"distance,":[63],"lacks":[64],"temporal":[65],"locality,":[66],"inflates":[68],"working":[69],"set":[70],"by":[71,145,172,178],"2-4\u00d7.":[72],"We":[73,86,92,112,186],"introduce":[74,93],"Tapeflow,":[75,161],"compiler":[77],"framework":[78],"orchestrate":[80,120],"manage":[82],"tape.":[85],"make":[87],"three":[88],"key":[89],"contributions.":[90],"i)":[91],"concept":[95],"of":[96,151,193],"regions,":[97],"which":[98],"transforms":[99],"layout":[102],"into":[103,116,148],"array-of-structs":[105],"format":[106],"improve":[108],"spatial":[109],"reuse.":[110],"ii)":[111],"schedule":[113],"execution":[115],"layers":[117],"explicitly":[119],"operands":[123,155],"using":[124],"scratchpad.":[126],"This":[127],"reduces":[128,174],"cache":[131],"size":[132],"on-chip":[134,175,184],"energy.":[135,185],"iii)":[136],"Finally,":[137],"we":[138],"stream":[139],"from":[142],"DRAM":[144],"organizing":[146],"FIFO":[150],"tiles.":[152],"arrive":[156],"just-in-time":[157],"each":[159],"layer.":[160],"running":[162],"on":[163,189],"same":[165],"hardware,":[166],"outperforms":[167],"Enzyme,":[168],"state-of-the-art":[170],"compiler,":[171],"1.3-2.5\u00d7,":[173],"SRAM":[176],"usage":[177],"5\u201340":[179],"\u00d7,":[180],"saves":[182],"8\u00d7":[183],"demonstrate":[187],"Tapeflow":[188],"wide":[191],"range":[192],"algorithms":[194],"written":[195],"general-purpose":[197],"language.":[198]},"counts_by_year":[{"year":2025,"cited_by_count":1},{"year":2024,"cited_by_count":1}],"updated_date":"2025-12-25T23:11:45.687758","created_date":"2025-10-10T00:00:00"}