{"id":"https://openalex.org/W4393931333","doi":"https://doi.org/10.1145/3642961.3643800","title":"Enhancing Intra-Node GPU-to-GPU Performance in MPI+UCX through Multi-Path Communication","display_name":"Enhancing Intra-Node GPU-to-GPU Performance in MPI+UCX through Multi-Path Communication","publication_year":2024,"publication_date":"2024-03-02","ids":{"openalex":"https://openalex.org/W4393931333","doi":"https://doi.org/10.1145/3642961.3643800"},"language":"en","primary_location":{"id":"doi:10.1145/3642961.3643800","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3642961.3643800","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3642961.3643800","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 3rd International Workshop on Extreme Heterogeneity Solutions","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3642961.3643800","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5031362631","display_name":"Amirhossein Sojoodi","orcid":"https://orcid.org/0000-0001-9877-3201"},"institutions":[{"id":"https://openalex.org/I204722609","display_name":"Queen's University","ror":"https://ror.org/02y72wh86","country_code":"CA","type":"education","lineage":["https://openalex.org/I204722609"]}],"countries":["CA"],"is_corresponding":true,"raw_author_name":"Amirhossein Sojoodi","raw_affiliation_strings":["Queen's University, Canada"],"raw_orcid":"https://orcid.org/0000-0001-9877-3201","affiliations":[{"raw_affiliation_string":"Queen's University, Canada","institution_ids":["https://openalex.org/I204722609"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5081527852","display_name":"Y\u0131ltan Hassan Temu\u00e7in","orcid":"https://orcid.org/0000-0002-4145-4848"},"institutions":[{"id":"https://openalex.org/I204722609","display_name":"Queen's University","ror":"https://ror.org/02y72wh86","country_code":"CA","type":"education","lineage":["https://openalex.org/I204722609"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Yiltan H. Temucin","raw_affiliation_strings":["Queen's University, Canada"],"raw_orcid":"https://orcid.org/0000-0002-4145-4848","affiliations":[{"raw_affiliation_string":"Queen's University, Canada","institution_ids":["https://openalex.org/I204722609"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5039854005","display_name":"Ahmad Afsahi","orcid":"https://orcid.org/0000-0002-2924-6851"},"institutions":[{"id":"https://openalex.org/I204722609","display_name":"Queen's University","ror":"https://ror.org/02y72wh86","country_code":"CA","type":"education","lineage":["https://openalex.org/I204722609"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Ahmad Afsahi","raw_affiliation_strings":["Queen's University, Canada"],"raw_orcid":"https://orcid.org/0000-0002-2924-6851","affiliations":[{"raw_affiliation_string":"Queen's University, Canada","institution_ids":["https://openalex.org/I204722609"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5031362631"],"corresponding_institution_ids":["https://openalex.org/I204722609"],"apc_list":null,"apc_paid":null,"fwci":1.4195,"has_fulltext":true,"cited_by_count":3,"citation_normalized_percentile":{"value":0.79542061,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":91,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"9","last_page":"14"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9991999864578247,"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":0.9991999864578247,"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/T11181","display_name":"Advanced Data Storage Technologies","score":0.9983999729156494,"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/T11478","display_name":"Caching and Content Delivery","score":0.9955000281333923,"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":0.9002853035926819},{"id":"https://openalex.org/keywords/pci-express","display_name":"PCI Express","score":0.7491626739501953},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.6858228445053101},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.6370803713798523},{"id":"https://openalex.org/keywords/node","display_name":"Node (physics)","score":0.5698318481445312},{"id":"https://openalex.org/keywords/cuda","display_name":"CUDA","score":0.5341504812240601},{"id":"https://openalex.org/keywords/host","display_name":"Host (biology)","score":0.514195442199707},{"id":"https://openalex.org/keywords/solver","display_name":"Solver","score":0.503149688243866},{"id":"https://openalex.org/keywords/path","display_name":"Path (computing)","score":0.45902734994888306},{"id":"https://openalex.org/keywords/message-passing","display_name":"Message passing","score":0.4198451638221741},{"id":"https://openalex.org/keywords/computer-network","display_name":"Computer network","score":0.195298969745636},{"id":"https://openalex.org/keywords/embedded-system","display_name":"Embedded system","score":0.16108283400535583},{"id":"https://openalex.org/keywords/field-programmable-gate-array","display_name":"Field-programmable gate array","score":0.09953087568283081}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.9002853035926819},{"id":"https://openalex.org/C64270927","wikidata":"https://www.wikidata.org/wiki/Q206924","display_name":"PCI Express","level":3,"score":0.7491626739501953},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.6858228445053101},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.6370803713798523},{"id":"https://openalex.org/C62611344","wikidata":"https://www.wikidata.org/wiki/Q1062658","display_name":"Node (physics)","level":2,"score":0.5698318481445312},{"id":"https://openalex.org/C2778119891","wikidata":"https://www.wikidata.org/wiki/Q477690","display_name":"CUDA","level":2,"score":0.5341504812240601},{"id":"https://openalex.org/C126831891","wikidata":"https://www.wikidata.org/wiki/Q221673","display_name":"Host (biology)","level":2,"score":0.514195442199707},{"id":"https://openalex.org/C2778770139","wikidata":"https://www.wikidata.org/wiki/Q1966904","display_name":"Solver","level":2,"score":0.503149688243866},{"id":"https://openalex.org/C2777735758","wikidata":"https://www.wikidata.org/wiki/Q817765","display_name":"Path (computing)","level":2,"score":0.45902734994888306},{"id":"https://openalex.org/C854659","wikidata":"https://www.wikidata.org/wiki/Q1859284","display_name":"Message passing","level":2,"score":0.4198451638221741},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.195298969745636},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.16108283400535583},{"id":"https://openalex.org/C42935608","wikidata":"https://www.wikidata.org/wiki/Q190411","display_name":"Field-programmable gate array","level":2,"score":0.09953087568283081},{"id":"https://openalex.org/C18903297","wikidata":"https://www.wikidata.org/wiki/Q7150","display_name":"Ecology","level":1,"score":0.0},{"id":"https://openalex.org/C66938386","wikidata":"https://www.wikidata.org/wiki/Q633538","display_name":"Structural engineering","level":1,"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/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.0},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3642961.3643800","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3642961.3643800","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3642961.3643800","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 3rd International Workshop on Extreme Heterogeneity Solutions","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3642961.3643800","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3642961.3643800","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3642961.3643800","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 3rd International Workshop on Extreme Heterogeneity Solutions","raw_type":"proceedings-article"},"sustainable_development_goals":[{"score":0.6600000262260437,"id":"https://metadata.un.org/sdg/7","display_name":"Affordable and clean energy"}],"awards":[],"funders":[{"id":"https://openalex.org/F4320331257","display_name":"Alliance de recherche num\u00e9rique du Canada","ror":"https://ror.org/010r6td27"},{"id":"https://openalex.org/F4320334593","display_name":"Natural Sciences and Engineering Research Council of Canada","ror":"https://ror.org/01h531d29"}],"has_content":{"grobid_xml":false,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4393931333.pdf"},"referenced_works_count":9,"referenced_works":["https://openalex.org/W1637731592","https://openalex.org/W1962931680","https://openalex.org/W2885707392","https://openalex.org/W2926767350","https://openalex.org/W3101708369","https://openalex.org/W3204371345","https://openalex.org/W4205832898","https://openalex.org/W4210863326","https://openalex.org/W4280577511"],"related_works":["https://openalex.org/W2058965144","https://openalex.org/W2164382479","https://openalex.org/W2146343568","https://openalex.org/W98480971","https://openalex.org/W2150291671","https://openalex.org/W2013643406","https://openalex.org/W2027972911","https://openalex.org/W2157978810","https://openalex.org/W2983282793","https://openalex.org/W1973046741"],"abstract_inverted_index":{"Efficient":[0],"communication":[1,19,28,61],"among":[2],"GPUs":[3],"is":[4],"crucial":[5],"for":[6],"achieving":[7,103],"high":[8],"performance":[9,29,52],"in":[10,65,97],"modern":[11],"GPU-accelerated":[12],"applications.":[13],"This":[14],"paper":[15],"introduces":[16],"a":[17,66],"multi-path":[18,69],"framework":[20],"within":[21],"the":[22,44,81,92,99],"MPI+UCX":[23],"library":[24],"to":[25,77,105],"enhance":[26],"P2P":[27,60,70],"between":[30],"intra-node":[31],"GPUs,":[32],"by":[33,55,75],"concurrently":[34],"leveraging":[35],"multiple":[36],"paths,":[37],"including":[38],"available":[39],"NVLinks":[40],"and":[41,84],"PCIe":[42],"through":[43],"host.":[45],"Through":[46],"extensive":[47],"experiments,":[48],"we":[49,90],"demonstrate":[50,91],"significant":[51],"gains":[53],"achieved":[54],"our":[56,95],"approach,":[57],"surpassing":[58],"baseline":[59],"methods.":[62],"More":[63],"specifically,":[64],"4-GPU":[67],"node,":[68],"improves":[71],"UCX":[72],"Put":[73],"bandwidth":[74],"up":[76,104],"2.85x":[78],"when":[79],"utilizing":[80],"host":[82],"path":[83],"2":[85],"other":[86],"GPU":[87],"paths.":[88],"Furthermore,":[89],"effectiveness":[93],"of":[94],"approach":[96],"accelerating":[98],"Jacobi":[100],"iterative":[101],"solver,":[102],"1.27x":[106],"runtime":[107],"speedup.":[108]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":1}],"updated_date":"2026-04-25T08:17:42.794288","created_date":"2025-10-10T00:00:00"}