{"id":"https://openalex.org/W4308091319","doi":"https://doi.org/10.1109/hpec55821.2022.9926341","title":"Analyzing Multi-trillion Edge Graphs on Large GPU Clusters: A Case Study with PageRank","display_name":"Analyzing Multi-trillion Edge Graphs on Large GPU Clusters: A Case Study with PageRank","publication_year":2022,"publication_date":"2022-09-19","ids":{"openalex":"https://openalex.org/W4308091319","doi":"https://doi.org/10.1109/hpec55821.2022.9926341"},"language":"en","primary_location":{"id":"doi:10.1109/hpec55821.2022.9926341","is_oa":false,"landing_page_url":"https://doi.org/10.1109/hpec55821.2022.9926341","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2022 IEEE High Performance Extreme Computing Conference (HPEC)","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/A5062348894","display_name":"Seunghwa Kang","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Seunghwa Kang","raw_affiliation_strings":["NVIDIA"],"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5041704325","display_name":"Joseph Nke","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Joseph Nke","raw_affiliation_strings":["NVIDIA"],"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5063365524","display_name":"Brad Rees","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Brad Rees","raw_affiliation_strings":["NVIDIA"],"affiliations":[{"raw_affiliation_string":"NVIDIA","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5062348894"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.8283,"has_fulltext":false,"cited_by_count":6,"citation_normalized_percentile":{"value":0.74342923,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":96},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"7"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11478","display_name":"Caching and Content Delivery","score":0.9976999759674072,"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"}},"topics":[{"id":"https://openalex.org/T11478","display_name":"Caching and Content Delivery","score":0.9976999759674072,"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"}},{"id":"https://openalex.org/T10064","display_name":"Complex Network Analysis Techniques","score":0.9876999855041504,"subfield":{"id":"https://openalex.org/subfields/3109","display_name":"Statistical and Nonlinear Physics"},"field":{"id":"https://openalex.org/fields/31","display_name":"Physics and Astronomy"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T12292","display_name":"Graph Theory and Algorithms","score":0.9855999946594238,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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":0.8643468022346497},{"id":"https://openalex.org/keywords/pagerank","display_name":"PageRank","score":0.8073489665985107},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.7655221223831177},{"id":"https://openalex.org/keywords/gpu-cluster","display_name":"GPU cluster","score":0.6473790407180786},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.5901473164558411},{"id":"https://openalex.org/keywords/supercomputer","display_name":"Supercomputer","score":0.568888783454895},{"id":"https://openalex.org/keywords/node","display_name":"Node (physics)","score":0.460714191198349},{"id":"https://openalex.org/keywords/column","display_name":"Column (typography)","score":0.4536045491695404},{"id":"https://openalex.org/keywords/cluster-analysis","display_name":"Cluster analysis","score":0.450662761926651},{"id":"https://openalex.org/keywords/vertex","display_name":"Vertex (graph theory)","score":0.4411846101284027},{"id":"https://openalex.org/keywords/infiniband","display_name":"InfiniBand","score":0.4368413984775543},{"id":"https://openalex.org/keywords/theoretical-computer-science","display_name":"Theoretical computer science","score":0.39159291982650757},{"id":"https://openalex.org/keywords/graph","display_name":"Graph","score":0.38525766134262085},{"id":"https://openalex.org/keywords/cuda","display_name":"CUDA","score":0.3734428286552429},{"id":"https://openalex.org/keywords/database","display_name":"Database","score":0.19501030445098877},{"id":"https://openalex.org/keywords/computer-network","display_name":"Computer network","score":0.1596980094909668},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.1069703996181488}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8643468022346497},{"id":"https://openalex.org/C2779172887","wikidata":"https://www.wikidata.org/wiki/Q184316","display_name":"PageRank","level":2,"score":0.8073489665985107},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.7655221223831177},{"id":"https://openalex.org/C2781335571","wikidata":"https://www.wikidata.org/wiki/Q2633544","display_name":"GPU cluster","level":3,"score":0.6473790407180786},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.5901473164558411},{"id":"https://openalex.org/C83283714","wikidata":"https://www.wikidata.org/wiki/Q121117","display_name":"Supercomputer","level":2,"score":0.568888783454895},{"id":"https://openalex.org/C62611344","wikidata":"https://www.wikidata.org/wiki/Q1062658","display_name":"Node (physics)","level":2,"score":0.460714191198349},{"id":"https://openalex.org/C2780551164","wikidata":"https://www.wikidata.org/wiki/Q2306599","display_name":"Column (typography)","level":3,"score":0.4536045491695404},{"id":"https://openalex.org/C73555534","wikidata":"https://www.wikidata.org/wiki/Q622825","display_name":"Cluster analysis","level":2,"score":0.450662761926651},{"id":"https://openalex.org/C80899671","wikidata":"https://www.wikidata.org/wiki/Q1304193","display_name":"Vertex (graph theory)","level":3,"score":0.4411846101284027},{"id":"https://openalex.org/C2781030343","wikidata":"https://www.wikidata.org/wiki/Q922437","display_name":"InfiniBand","level":2,"score":0.4368413984775543},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.39159291982650757},{"id":"https://openalex.org/C132525143","wikidata":"https://www.wikidata.org/wiki/Q141488","display_name":"Graph","level":2,"score":0.38525766134262085},{"id":"https://openalex.org/C2778119891","wikidata":"https://www.wikidata.org/wiki/Q477690","display_name":"CUDA","level":2,"score":0.3734428286552429},{"id":"https://openalex.org/C77088390","wikidata":"https://www.wikidata.org/wiki/Q8513","display_name":"Database","level":1,"score":0.19501030445098877},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.1596980094909668},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.1069703996181488},{"id":"https://openalex.org/C126042441","wikidata":"https://www.wikidata.org/wiki/Q1324888","display_name":"Frame (networking)","level":2,"score":0.0},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.0},{"id":"https://openalex.org/C66938386","wikidata":"https://www.wikidata.org/wiki/Q633538","display_name":"Structural engineering","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/hpec55821.2022.9926341","is_oa":false,"landing_page_url":"https://doi.org/10.1109/hpec55821.2022.9926341","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2022 IEEE High Performance Extreme Computing Conference (HPEC)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W1597650818","https://openalex.org/W2056717482","https://openalex.org/W3101208208","https://openalex.org/W2068528924","https://openalex.org/W1966243865","https://openalex.org/W1449833061","https://openalex.org/W2017587301","https://openalex.org/W619618554","https://openalex.org/W2049347805","https://openalex.org/W2030707850"],"abstract_inverted_index":{"We":[0,59,199,307],"previously":[1],"reported":[2],"PageRank":[3,113],"performance":[4],"results":[5],"on":[6,204,259,300],"a":[7,71,122,151,269],"cluster":[8,187],"with":[9,64,76,112,197,304],"32":[10,185,243,261,285],"A100":[11,189,224],"GPUs":[12,56,80,190,225],"[7].":[13],"This":[14],"paper":[15,310],"extends":[16],"the":[17,31,41,44,53,65,95,98,110,119,141,180,201,205,232,260,301,318],"previous":[18,24,62],"work":[19],"to":[20,40,82,97,171,316],"2048":[21,305],"GPUs.":[22,306],"The":[23,184],"implementation":[25,63,203],"performs":[26],"well":[27],"as":[28,30,52],"long":[29],"number":[32,54],"of":[33,43,55,121,125,268],"G":[34],"PU":[35],"s":[36],"is":[37,150],"small":[38],"relative":[39],"square":[42],"average":[45,248,289],"vertex":[46,175,245,249,283,290],"degree":[47],"but":[48],"its":[49],"scalability":[50,120],"deteriorates":[51],"further":[57],"increases.":[58],"updated":[60,134,202],"our":[61,135],"following":[66],"objectives:":[67],"1)":[68],"enable":[69],"analyzing":[70],"P":[72,77,83],"times":[73,78],"larger":[74],"graph":[75,235,273],"more":[79,131],"up":[81],"=":[84],"2048,":[85],"2)":[86],"achieve":[87],"reasonably":[88],"good":[89],"weak":[90],"scaling,":[91],"and":[92,161,178,214,239,277],"3)":[93],"integrate":[94],"improvements":[96],"open-source":[99],"data":[100,153],"science":[101],"ecosystem":[102],"(i.e.":[103],"RAPIDS":[104],"cuGraph,":[105],"https://github.com/rapidsai/cugraph).":[106],"While":[107],"we":[108,133],"evaluate":[109],"updates":[111],"in":[114,127],"this":[115,309],"paper,":[116],"they":[117],"improve":[118,317],"broader":[123],"set":[124],"algorithms":[126],"cuGraph.":[128],"To":[129],"be":[130],"specific,":[132],"2D":[136],"edge":[137,173,287],"partitioning":[138],"scheme;":[139],"implemented":[140],"PDCSC":[142],"(partially":[143],"doubly":[144],"compressed":[145,164],"sparse":[146,159,165],"column)":[147,160],"format":[148],"which":[149,208],"hybrid":[152],"structure":[154],"that":[155],"combines":[156],"CSC":[157],"(compressed":[158],"DCSC":[162],"(doubly":[163],"column);":[166],"adopted":[167],"(key,":[168],"value)":[169],"pairs":[170],"store":[172],"source":[174],"property":[176],"values;":[177],"improved":[179],"reduction":[181],"communication":[182,213],"strategy.":[183],"GPU":[186,262],"has":[188,222],"(40":[191],"GB":[192,227],"HBM":[193,228],"per":[194,229,255,296],"GPU)":[195],"connected":[196],"NVLink.":[198],"ran":[200],"Selene":[206,220,302],"supercomputer":[207,303],"uses":[209],"InfiniBand":[210],"for":[211,216],"inter-node":[212],"NVLink":[215],"intra-node":[217],"communication.":[218],"Each":[219],"node":[221],"eight":[223],"(80":[226],"GPU).":[230],"Analyzing":[231],"web":[233],"crawl":[234],"(3.563":[236],"billion":[237,241,275],"vertices":[238,276],"128.7":[240],"edges,":[242,280],"bit":[244,282,286],"ID,":[246,284],"unweighted,":[247],"degree:":[250,291],"36.12)":[251],"took":[252,293],"0.187":[253],"second":[254,295],"Page":[256,265,297],"Rank":[257,266,298],"iteration":[258,299],"cluster.":[263],"Computing":[264],"scores":[267],"scale":[270],"38":[271],"R-mat":[272],"(274.9":[274],"4.398":[278],"trillion":[279],"64":[281],"weight,":[288],"16)":[292],"1.54":[294],"conclude":[308],"discussing":[311],"potential":[312],"network":[313],"system":[314],"enhancements":[315],"scaling.":[319]},"counts_by_year":[{"year":2025,"cited_by_count":2},{"year":2024,"cited_by_count":2},{"year":2023,"cited_by_count":2}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}