{"id":"https://openalex.org/W4409248543","doi":"https://doi.org/10.1109/hpca61900.2025.00114","title":"TidalMesh: Topology-Driven AllReduce Collective Communication for Mesh Topology","display_name":"TidalMesh: Topology-Driven AllReduce Collective Communication for Mesh Topology","publication_year":2025,"publication_date":"2025-03-01","ids":{"openalex":"https://openalex.org/W4409248543","doi":"https://doi.org/10.1109/hpca61900.2025.00114"},"language":"en","primary_location":{"id":"doi:10.1109/hpca61900.2025.00114","is_oa":false,"landing_page_url":"https://doi.org/10.1109/hpca61900.2025.00114","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 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/A5021761169","display_name":"Dongkyun Lim","orcid":null},"institutions":[{"id":"https://openalex.org/I4210099236","display_name":"Kootenay Association for Science & Technology","ror":"https://ror.org/011pv9p44","country_code":"CA","type":"nonprofit","lineage":["https://openalex.org/I4210099236"]},{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["CA","KR"],"is_corresponding":true,"raw_author_name":"Dongkyun Lim","raw_affiliation_strings":["KAIST"],"affiliations":[{"raw_affiliation_string":"KAIST","institution_ids":["https://openalex.org/I4210099236","https://openalex.org/I157485424"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100698124","display_name":"John Kim","orcid":"https://orcid.org/0000-0003-3958-3891"},"institutions":[{"id":"https://openalex.org/I4210099236","display_name":"Kootenay Association for Science & Technology","ror":"https://ror.org/011pv9p44","country_code":"CA","type":"nonprofit","lineage":["https://openalex.org/I4210099236"]},{"id":"https://openalex.org/I157485424","display_name":"Korea Advanced Institute of Science and Technology","ror":"https://ror.org/05apxxy63","country_code":"KR","type":"education","lineage":["https://openalex.org/I157485424"]}],"countries":["CA","KR"],"is_corresponding":false,"raw_author_name":"John Kim","raw_affiliation_strings":["KAIST"],"affiliations":[{"raw_affiliation_string":"KAIST","institution_ids":["https://openalex.org/I4210099236","https://openalex.org/I157485424"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5021761169"],"corresponding_institution_ids":["https://openalex.org/I157485424","https://openalex.org/I4210099236"],"apc_list":null,"apc_paid":null,"fwci":4.0327,"has_fulltext":false,"cited_by_count":5,"citation_normalized_percentile":{"value":0.93347723,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":96,"max":100},"biblio":{"volume":null,"issue":null,"first_page":"1526","last_page":"1540"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12784","display_name":"Modular Robots and Swarm Intelligence","score":0.965499997138977,"subfield":{"id":"https://openalex.org/subfields/2210","display_name":"Mechanical Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T12784","display_name":"Modular Robots and Swarm Intelligence","score":0.965499997138977,"subfield":{"id":"https://openalex.org/subfields/2210","display_name":"Mechanical 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/T10789","display_name":"Interactive and Immersive Displays","score":0.9240999817848206,"subfield":{"id":"https://openalex.org/subfields/1709","display_name":"Human-Computer Interaction"},"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/topology","display_name":"Topology (electrical circuits)","score":0.7269145250320435},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.5704448819160461},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.13817444443702698},{"id":"https://openalex.org/keywords/electrical-engineering","display_name":"Electrical engineering","score":0.06288599967956543}],"concepts":[{"id":"https://openalex.org/C184720557","wikidata":"https://www.wikidata.org/wiki/Q7825049","display_name":"Topology (electrical circuits)","level":2,"score":0.7269145250320435},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5704448819160461},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.13817444443702698},{"id":"https://openalex.org/C119599485","wikidata":"https://www.wikidata.org/wiki/Q43035","display_name":"Electrical engineering","level":1,"score":0.06288599967956543}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/hpca61900.2025.00114","is_oa":false,"landing_page_url":"https://doi.org/10.1109/hpca61900.2025.00114","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE International Symposium on High Performance Computer Architecture (HPCA)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/10","score":0.5400000214576721,"display_name":"Reduced inequalities"}],"awards":[],"funders":[{"id":"https://openalex.org/F4320332195","display_name":"Samsung","ror":"https://ror.org/04w3jy968"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":48,"referenced_works":["https://openalex.org/W1536680647","https://openalex.org/W2057332538","https://openalex.org/W2061319431","https://openalex.org/W2097117768","https://openalex.org/W2118231264","https://openalex.org/W2127533540","https://openalex.org/W2131613942","https://openalex.org/W2136041296","https://openalex.org/W2164945803","https://openalex.org/W2194775991","https://openalex.org/W2605350416","https://openalex.org/W2618530766","https://openalex.org/W2766447205","https://openalex.org/W2955454939","https://openalex.org/W2980104813","https://openalex.org/W2981114289","https://openalex.org/W3047357290","https://openalex.org/W3086105743","https://openalex.org/W3089731223","https://openalex.org/W3092260380","https://openalex.org/W3097528158","https://openalex.org/W3190806564","https://openalex.org/W4232919122","https://openalex.org/W4281790033","https://openalex.org/W4297097375","https://openalex.org/W4352977393","https://openalex.org/W4360764808","https://openalex.org/W4360831831","https://openalex.org/W4376130831","https://openalex.org/W4380874786","https://openalex.org/W4393406935","https://openalex.org/W4394998995","https://openalex.org/W4395117348","https://openalex.org/W4402044182","https://openalex.org/W4404955085","https://openalex.org/W6739901393","https://openalex.org/W6748645090","https://openalex.org/W6756439839","https://openalex.org/W6758823024","https://openalex.org/W6766978945","https://openalex.org/W6767997687","https://openalex.org/W6769475105","https://openalex.org/W6786394717","https://openalex.org/W6805453576","https://openalex.org/W6839137647","https://openalex.org/W6853227453","https://openalex.org/W6864930665","https://openalex.org/W6892368343"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052"],"abstract_inverted_index":{"In":[0,92,127,194],"deep":[1,240],"learning":[2,241],"workloads,":[3],"collective":[4,28,111,199,210],"communication":[5,29,112,200,211],"across":[6,238],"multiple":[7],"nodes":[8,62],"is":[9,24,133],"a":[10,25,88,97,108,146,169,224,244,261],"critical":[11],"component":[12],"in":[13,136,192,228,235],"determining":[14],"overall":[15],"performance.":[16,193],"AllReduce":[17,40,71,83,110,143,183,236],"(as":[18],"well":[19,51,122],"as":[20,50,52,129],"ReduceScatter":[21],"and":[22],"AllGather)":[23],"commonly":[26],"used":[27,135],"for":[30,87,176],"not":[31],"only":[32],"training":[33],"but":[34,73],"also":[35],"inference.":[36],"The":[37],"performance":[38,72,175,237],"of":[39,56,118,154,187,209,217,221,260],"depends":[41],"on":[42,69,243],"the":[43,47,53,57,61,101,115,119,124,151,155,164,214,254,258],"algorithm":[44,120,144],"utilized":[45],"(or":[46],"\u201clogical\u201d":[48],"topology)":[49],"physical":[54,90,125,156],"topology":[55,102,117,132,159],"system":[58],"that":[59,149],"interconnects":[60],"together.":[63],"There":[64],"has":[65],"been":[66,78],"many":[67],"work":[68,75],"improving":[70],"prior":[74],"have":[76],"often":[77],"topologyaware":[79],"approach":[80,99,148],"where":[81,100],"existing":[82],"algorithms":[84],"were":[85],"optimized":[86],"given":[89],"topology.":[91,126],"this":[93],"work,":[94],"we":[95,140,196],"propose":[96,107,141,179],"topology-driven":[98],"characteristics":[103,153],"are":[104],"exploited":[105,204],"to":[106,123,168,172,205,230,253],"novel":[109,147],"algorithm;":[113],"thus,":[114],"logical":[116,262],"maps":[121],"particular,":[128],"2D":[130,157,250],"mesh":[131,158],"widely":[134],"various":[137,239],"scale-out":[138],"systems,":[139],"TidalMesh":[142,182,188,226],"-":[145],"exploits":[150],"inherent":[152],"by":[160],"pushing":[161],"flows":[162],"between":[163],"endpoint":[165],"nodes,":[166],"similar":[167],"tidal":[170],"wave,":[171],"achieve":[173],"near-optimal":[174],"AllReduce.":[177,222],"We":[178],"how":[180,198],"Sparse":[181],"minimizes":[184],"bandwidth":[185],"overhead":[186],"with":[189],"no":[190],"loss":[191],"addition,":[195],"demonstrate":[197],"unrolling":[201],"can":[202],"be":[203],"enable":[206],"\u201csoftware":[207],"pipelining\u201d":[208],"while":[212,256],"exploiting":[213],"unique":[215],"opportunity":[216],"superimposing":[218],"different":[219],"phases":[220],"As":[223],"result,":[225],"results":[227],"up":[229],"$\\mathbf{2":[231],"4":[232],"\\%}$":[233],"improvement":[234],"models":[242],"64":[245],"-node":[246],"$8":[247],"\\times":[248],"8$":[249],"mesh,":[251],"compared":[252],"state-of-the-art":[255],"maintaining":[257],"simplicity":[259],"ring":[263],"algorithm.":[264]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":3}],"updated_date":"2026-03-06T13:50:29.536080","created_date":"2025-10-10T00:00:00"}