{"id":"https://openalex.org/W3141076774","doi":"https://doi.org/10.1109/hpca51647.2021.00015","title":"A Computational Stack for Cross-Domain Acceleration","display_name":"A Computational Stack for Cross-Domain Acceleration","publication_year":2021,"publication_date":"2021-02-01","ids":{"openalex":"https://openalex.org/W3141076774","doi":"https://doi.org/10.1109/hpca51647.2021.00015","mag":"3141076774"},"language":"en","primary_location":{"id":"doi:10.1109/hpca51647.2021.00015","is_oa":false,"landing_page_url":"https://doi.org/10.1109/hpca51647.2021.00015","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2021 IEEE International Symposium on High-Performance Computer Architecture (HPCA)","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/A5000799803","display_name":"Sean Kinzer","orcid":"https://orcid.org/0000-0002-0955-585X"},"institutions":[{"id":"https://openalex.org/I36258959","display_name":"University of California, San Diego","ror":"https://ror.org/0168r3w48","country_code":"US","type":"education","lineage":["https://openalex.org/I36258959"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Sean Kinzer","raw_affiliation_strings":["Alternative Computing Technologies (ACT) Lab, University of California San Diego"],"affiliations":[{"raw_affiliation_string":"Alternative Computing Technologies (ACT) Lab, University of California San Diego","institution_ids":["https://openalex.org/I36258959"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5018716691","display_name":"Joon Kyung Kim","orcid":"https://orcid.org/0000-0002-7984-9669"},"institutions":[{"id":"https://openalex.org/I36258959","display_name":"University of California, San Diego","ror":"https://ror.org/0168r3w48","country_code":"US","type":"education","lineage":["https://openalex.org/I36258959"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Joon Kyung Kim","raw_affiliation_strings":["Alternative Computing Technologies (ACT) Lab, University of California San Diego"],"affiliations":[{"raw_affiliation_string":"Alternative Computing Technologies (ACT) Lab, University of California San Diego","institution_ids":["https://openalex.org/I36258959"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5082499242","display_name":"Soroush Ghodrati","orcid":"https://orcid.org/0000-0001-5514-8027"},"institutions":[{"id":"https://openalex.org/I36258959","display_name":"University of California, San Diego","ror":"https://ror.org/0168r3w48","country_code":"US","type":"education","lineage":["https://openalex.org/I36258959"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Soroush Ghodrati","raw_affiliation_strings":["Alternative Computing Technologies (ACT) Lab, University of California San Diego"],"affiliations":[{"raw_affiliation_string":"Alternative Computing Technologies (ACT) Lab, University of California San Diego","institution_ids":["https://openalex.org/I36258959"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5076569616","display_name":"Brahmendra Reddy Yatham","orcid":null},"institutions":[{"id":"https://openalex.org/I36258959","display_name":"University of California, San Diego","ror":"https://ror.org/0168r3w48","country_code":"US","type":"education","lineage":["https://openalex.org/I36258959"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Brahmendra Yatham","raw_affiliation_strings":["Alternative Computing Technologies (ACT) Lab, University of California San Diego"],"affiliations":[{"raw_affiliation_string":"Alternative Computing Technologies (ACT) Lab, University of California San Diego","institution_ids":["https://openalex.org/I36258959"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5012052294","display_name":"Alric Althoff","orcid":"https://orcid.org/0000-0002-8355-6906"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Alric Althoff","raw_affiliation_strings":["Tortuga Logic"],"affiliations":[{"raw_affiliation_string":"Tortuga Logic","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5089590312","display_name":"Divya Mahajan","orcid":"https://orcid.org/0009-0007-8184-0528"},"institutions":[{"id":"https://openalex.org/I4210164937","display_name":"Microsoft Research (United Kingdom)","ror":"https://ror.org/05k87vq12","country_code":"GB","type":"company","lineage":["https://openalex.org/I1290206253","https://openalex.org/I4210164937"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Divya Mahajan","raw_affiliation_strings":["Microsoft"],"affiliations":[{"raw_affiliation_string":"Microsoft","institution_ids":["https://openalex.org/I4210164937"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5043665959","display_name":"Sorin Lerner","orcid":"https://orcid.org/0000-0003-3957-0628"},"institutions":[{"id":"https://openalex.org/I36258959","display_name":"University of California, San Diego","ror":"https://ror.org/0168r3w48","country_code":"US","type":"education","lineage":["https://openalex.org/I36258959"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Sorin Lerner","raw_affiliation_strings":["Alternative Computing Technologies (ACT) Lab, University of California San Diego"],"affiliations":[{"raw_affiliation_string":"Alternative Computing Technologies (ACT) Lab, University of California San Diego","institution_ids":["https://openalex.org/I36258959"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5084514143","display_name":"Hadi Esmaeilzadeh","orcid":"https://orcid.org/0000-0002-8548-1039"},"institutions":[{"id":"https://openalex.org/I36258959","display_name":"University of California, San Diego","ror":"https://ror.org/0168r3w48","country_code":"US","type":"education","lineage":["https://openalex.org/I36258959"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Hadi Esmaeilzadeh","raw_affiliation_strings":["Alternative Computing Technologies (ACT) Lab, University of California San Diego"],"affiliations":[{"raw_affiliation_string":"Alternative Computing Technologies (ACT) Lab, University of California San Diego","institution_ids":["https://openalex.org/I36258959"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":8,"corresponding_author_ids":["https://openalex.org/A5000799803"],"corresponding_institution_ids":["https://openalex.org/I36258959"],"apc_list":null,"apc_paid":null,"fwci":1.3817,"has_fulltext":false,"cited_by_count":7,"citation_normalized_percentile":{"value":0.7903124,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":96},"biblio":{"volume":null,"issue":null,"first_page":"54","last_page":"70"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9994000196456909,"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":0.9994000196456909,"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/T12808","display_name":"Ferroelectric and Negative Capacitance Devices","score":0.9977999925613403,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10743","display_name":"Software Testing and Debugging Techniques","score":0.9933000206947327,"subfield":{"id":"https://openalex.org/subfields/1712","display_name":"Software"},"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":0.7980008125305176},{"id":"https://openalex.org/keywords/domain","display_name":"Domain (mathematical analysis)","score":0.6056407690048218},{"id":"https://openalex.org/keywords/modular-design","display_name":"Modular design","score":0.5448929071426392},{"id":"https://openalex.org/keywords/domain-specific-language","display_name":"Domain-specific language","score":0.5301526188850403},{"id":"https://openalex.org/keywords/granularity","display_name":"Granularity","score":0.4904050827026367},{"id":"https://openalex.org/keywords/theoretical-computer-science","display_name":"Theoretical computer science","score":0.47595998644828796},{"id":"https://openalex.org/keywords/analytics","display_name":"Analytics","score":0.43851327896118164},{"id":"https://openalex.org/keywords/polymath","display_name":"Polymath","score":0.4165080785751343},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.37111103534698486},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.34252020716667175},{"id":"https://openalex.org/keywords/programming-language","display_name":"Programming language","score":0.2926677465438843},{"id":"https://openalex.org/keywords/database","display_name":"Database","score":0.16840240359306335}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7980008125305176},{"id":"https://openalex.org/C36503486","wikidata":"https://www.wikidata.org/wiki/Q11235244","display_name":"Domain (mathematical analysis)","level":2,"score":0.6056407690048218},{"id":"https://openalex.org/C101468663","wikidata":"https://www.wikidata.org/wiki/Q1620158","display_name":"Modular design","level":2,"score":0.5448929071426392},{"id":"https://openalex.org/C135257023","wikidata":"https://www.wikidata.org/wiki/Q691358","display_name":"Domain-specific language","level":2,"score":0.5301526188850403},{"id":"https://openalex.org/C177774035","wikidata":"https://www.wikidata.org/wiki/Q1246948","display_name":"Granularity","level":2,"score":0.4904050827026367},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.47595998644828796},{"id":"https://openalex.org/C79158427","wikidata":"https://www.wikidata.org/wiki/Q485396","display_name":"Analytics","level":2,"score":0.43851327896118164},{"id":"https://openalex.org/C2776603922","wikidata":"https://www.wikidata.org/wiki/Q270141","display_name":"Polymath","level":2,"score":0.4165080785751343},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.37111103534698486},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.34252020716667175},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.2926677465438843},{"id":"https://openalex.org/C77088390","wikidata":"https://www.wikidata.org/wiki/Q8513","display_name":"Database","level":1,"score":0.16840240359306335},{"id":"https://openalex.org/C95457728","wikidata":"https://www.wikidata.org/wiki/Q309","display_name":"History","level":0,"score":0.0},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.0},{"id":"https://openalex.org/C166957645","wikidata":"https://www.wikidata.org/wiki/Q23498","display_name":"Archaeology","level":1,"score":0.0},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/hpca51647.2021.00015","is_oa":false,"landing_page_url":"https://doi.org/10.1109/hpca51647.2021.00015","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2021 IEEE International Symposium on High-Performance Computer Architecture (HPCA)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"score":0.7699999809265137,"display_name":"Affordable and clean energy","id":"https://metadata.un.org/sdg/7"}],"awards":[],"funders":[{"id":"https://openalex.org/F4320307764","display_name":"Microsoft","ror":"https://ror.org/00d0nc645"},{"id":"https://openalex.org/F4320308258","display_name":"Qualcomm","ror":"https://ror.org/002zrf773"},{"id":"https://openalex.org/F4320309327","display_name":"Google","ror":"https://ror.org/00njsd438"},{"id":"https://openalex.org/F4320314786","display_name":"Xilinx","ror":"https://ror.org/01rb7bk56"},{"id":"https://openalex.org/F4320315952","display_name":"Leidos","ror":"https://ror.org/012cvds63"},{"id":"https://openalex.org/F4320338279","display_name":"Air Force Office of Scientific Research","ror":"https://ror.org/011e9bt93"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":89,"referenced_works":["https://openalex.org/W75509891","https://openalex.org/W1668251704","https://openalex.org/W1783256592","https://openalex.org/W1919002818","https://openalex.org/W1964471912","https://openalex.org/W1974967412","https://openalex.org/W1977164425","https://openalex.org/W1983157164","https://openalex.org/W2006312753","https://openalex.org/W2035080386","https://openalex.org/W2042545849","https://openalex.org/W2048266589","https://openalex.org/W2050918004","https://openalex.org/W2052308318","https://openalex.org/W2054625910","https://openalex.org/W2065439108","https://openalex.org/W2098335003","https://openalex.org/W2101196063","https://openalex.org/W2102182691","https://openalex.org/W2105254232","https://openalex.org/W2112796928","https://openalex.org/W2142120580","https://openalex.org/W2146292423","https://openalex.org/W2149684715","https://openalex.org/W2153185479","https://openalex.org/W2153331583","https://openalex.org/W2162385899","https://openalex.org/W2165693128","https://openalex.org/W2171927842","https://openalex.org/W2194775991","https://openalex.org/W2240938131","https://openalex.org/W2261808795","https://openalex.org/W2271840356","https://openalex.org/W2411785910","https://openalex.org/W2412589610","https://openalex.org/W2442974303","https://openalex.org/W2562773490","https://openalex.org/W2566870951","https://openalex.org/W2567660513","https://openalex.org/W2606722458","https://openalex.org/W2612445135","https://openalex.org/W2780077279","https://openalex.org/W2786320458","https://openalex.org/W2786332448","https://openalex.org/W2789784730","https://openalex.org/W2798341898","https://openalex.org/W2798482398","https://openalex.org/W2798956872","https://openalex.org/W2804032941","https://openalex.org/W2805566098","https://openalex.org/W2868091835","https://openalex.org/W2883350506","https://openalex.org/W2883929540","https://openalex.org/W2889424113","https://openalex.org/W2899771611","https://openalex.org/W2943963155","https://openalex.org/W2963114857","https://openalex.org/W2995723163","https://openalex.org/W3001542287","https://openalex.org/W3037521362","https://openalex.org/W3041178375","https://openalex.org/W3090198137","https://openalex.org/W3098220359","https://openalex.org/W3101026687","https://openalex.org/W3120740533","https://openalex.org/W3148943889","https://openalex.org/W4236433846","https://openalex.org/W4241140669","https://openalex.org/W4242577057","https://openalex.org/W4245199738","https://openalex.org/W4246166885","https://openalex.org/W4255097844","https://openalex.org/W4297775537","https://openalex.org/W4297803378","https://openalex.org/W6603060583","https://openalex.org/W6651700774","https://openalex.org/W6662587704","https://openalex.org/W6676070312","https://openalex.org/W6684528527","https://openalex.org/W6690291838","https://openalex.org/W6694517276","https://openalex.org/W6737664043","https://openalex.org/W6748324931","https://openalex.org/W6751109714","https://openalex.org/W6753048956","https://openalex.org/W6756040250","https://openalex.org/W6772040440","https://openalex.org/W6774753494","https://openalex.org/W6785733332"],"related_works":["https://openalex.org/W4210325698","https://openalex.org/W2477016040","https://openalex.org/W2497643139","https://openalex.org/W2189139778","https://openalex.org/W4390846453","https://openalex.org/W4402226951","https://openalex.org/W2744356630","https://openalex.org/W3152533027","https://openalex.org/W2931688134","https://openalex.org/W2108960318"],"abstract_inverted_index":{"Domain-specific":[0],"accelerators":[1,11,232],"obtain":[2],"performance":[3,82,278,296],"benefits":[4],"by":[5,83,129,224,280],"restricting":[6],"their":[7],"algorithmic":[8],"domain.":[9,43],"These":[10],"utilize":[12],"specialized":[13],"languages":[14],"constrained":[15],"to":[16,38,71,110,137,177,193,230,237,256,288,343,375,389],"particular":[17],"hardware,":[18],"thus":[19,186],"trading":[20],"off":[21],"expressiveness":[22,80,276,359],"for":[23,33,183,191,204,297],"high":[24],"performance.":[25],"The":[26,44],"pendulum":[27],"has":[28],"swung":[29],"from":[30,313,378],"one":[31,39,346],"hardware":[32,40],"all":[34,138,178,320],"domains":[35,213,236,381],"(general-purpose":[36],"processors)":[37],"per":[41],"individual":[42],"middle-ground":[45],"on":[46,368],"this":[47,73],"spectrum-which":[48],"provides":[49],"a":[50,85,93,102,130,159,245,363],"unified":[51],"computational":[52],"stack":[53,91,201],"across":[54,113,197,233],"multiple,":[55],"but":[56],"not":[57],"all,":[58],"domains-":[59],"is":[60,127,228],"an":[61,188,239,323],"emerging":[62],"and":[63,75,81,104,123,218,259,266,277,291,300,336,358],"open":[64],"research":[65],"challenge.":[66],"This":[67,90],"paper":[68],"sets":[69],"out":[70],"explore":[72],"region":[74],"its":[76,285],"associated":[77],"tradeoff":[78,279],"between":[79],"defining":[84],"cross-domain":[86,95,262,275],"stack,":[87],"dubbed":[88],"PolyMath.":[89],"defines":[92],"high-level":[94],"language":[96],"(CDL),":[97],"called":[98,143],"PMLang,":[99],"that":[100,156,207,223,318],"in":[101,153,252,269,331],"modular":[103],"reusable":[105],"manner":[106],"encapsulates":[107],"mathematical":[108],"properties":[109],"be":[111],"expressive":[112],"multiple":[114,198,212,314],"domains-Robotics,":[115],"Graph":[116],"Analytics,":[117],"Digital":[118],"Signal":[119],"Processing,":[120],"Deep":[121],"Learning,":[122],"Data":[124],"Analytics.":[125],"PMLang":[126],"backed":[128],"recursively-defined":[131],"intermediate":[132],"representation":[133],"allowing":[134],"simultaneous":[135,175],"access":[136,176],"levels":[139],"of":[140,161,170,181,211,242,293,308,360,386],"operation":[141],"granularity,":[142],"sr":[144,172],"DFG.":[145],"Accelerator-specific":[146],"or":[147],"domain-specific":[148],"IRs":[149,196],"commonly":[150],"capture":[151],"operations":[152],"the":[154,167,171,179,209,234,274,294,304,345,356],"granularity":[155],"best":[157],"fits":[158],"set":[160],"Domain-Specific":[162],"Architectures":[163],"(DSAs).":[164],"In":[165,254],"contrast,":[166],"recursive":[168],"nature":[169],"DFG":[173],"enables":[174],"granularities":[180],"computation":[182],"every":[184],"operation,":[185],"forming":[187],"ideal":[189],"bridge":[190],"converting":[192],"various":[194],"DSA-specific":[195],"domains.":[199],"Our":[200],"unlocks":[202],"multi-acceleration":[203],"end-to-end":[205,301,309],"applications":[206,310],"cross":[208],"boundary":[210],"each":[214,282],"comprising":[215],"different":[216,380],"data":[217],"compute":[219],"patterns.":[220],"Evaluations":[221],"show":[222,317],"using":[225],"PolyMath":[226,361,370],"it":[227],"possible":[229],"harness":[231],"five":[235],"realize":[238],"average":[240],"speedup":[241,326,338],"3.3\u00d7":[243],"over":[244,327,333,339],"Xeon":[246],"CPU":[247],"along":[248],"with":[249,382],"18.1\u00d7":[250],"reduction":[251],"energy.":[253],"comparison":[255],"Jetson":[257,340],"Xavier":[258,341],"Titan":[260,334],"XP,":[261],"acceleration":[263],"offers":[264,322],"1.7\u00d7":[265],"7.2\u00d7":[267],"improvement":[268,330],"performance-per-watt,":[270],"respectively.":[271],"We":[272],"measure":[273],"comparing":[281],"benchmark":[283],"against":[284],"hand-optimized":[286],"implementation":[287],"achieve":[289],"83.9%":[290],"76.8%":[292],"optimal":[295],"single-domain":[298,349],"algorithms":[299,312,377],"applications.":[302],"For":[303],"two":[305,379],"case":[306],"studies":[307],"(comprising":[311],"domains),":[315],"results":[316],"accelerating":[319],"kernels":[321],"additional":[324],"2.0\u00d7":[325],"CPU,":[328],"6.1\u00d7":[329],"performance-per-watt":[332],"Xp,":[335],"2.8\u00d7":[337],"compared":[342],"only":[344],"most":[347],"effective":[348],"kernel":[350],"being":[351],"accelerated.":[352],"Finally,":[353],"we":[354],"examine":[355],"utility":[357],"through":[362],"user":[364],"study,":[365],"which":[366],"shows,":[367],"average,":[369],"requires":[371],"1.9\u00d7":[372],"less":[373],"time":[374],"implement":[376],"2.5\u00d7":[383],"fewer":[384],"lines":[385],"code":[387],"relative":[388],"Python.":[390]},"counts_by_year":[{"year":2025,"cited_by_count":1},{"year":2024,"cited_by_count":1},{"year":2023,"cited_by_count":2},{"year":2022,"cited_by_count":2},{"year":2021,"cited_by_count":1}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}