{"id":"https://openalex.org/W7155498797","doi":"https://doi.org/10.1145/3767295.3803620","title":"RoPeerTo: A Datacenter-Scale Architecture for Peer-To-Peer DMA between GPUs and FPGAs","display_name":"RoPeerTo: A Datacenter-Scale Architecture for Peer-To-Peer DMA between GPUs and FPGAs","publication_year":2026,"publication_date":"2026-04-24","ids":{"openalex":"https://openalex.org/W7155498797","doi":"https://doi.org/10.1145/3767295.3803620"},"language":"en","primary_location":{"id":"doi:10.1145/3767295.3803620","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3767295.3803620","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 21st European Conference on Computer Systems","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3767295.3803620","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5085174139","display_name":"M. V\u00e9nere","orcid":"https://orcid.org/0009-0002-8991-1443"},"institutions":[{"id":"https://openalex.org/I93860229","display_name":"Politecnico di Milano","ror":"https://ror.org/01nffqt88","country_code":"IT","type":"education","lineage":["https://openalex.org/I93860229"]}],"countries":["IT"],"is_corresponding":false,"raw_author_name":"Marco Venere","raw_affiliation_strings":["Politecnico di Milano, Milan, Italy"],"raw_orcid":"https://orcid.org/0009-0002-8991-1443","affiliations":[{"raw_affiliation_string":"Politecnico di Milano, Milan, Italy","institution_ids":["https://openalex.org/I93860229"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103099232","display_name":"Giuseppe Sorrentino","orcid":"https://orcid.org/0009-0009-3025-8620"},"institutions":[{"id":"https://openalex.org/I93860229","display_name":"Politecnico di Milano","ror":"https://ror.org/01nffqt88","country_code":"IT","type":"education","lineage":["https://openalex.org/I93860229"]}],"countries":["IT"],"is_corresponding":false,"raw_author_name":"Giuseppe Sorrentino","raw_affiliation_strings":["Politecnico di Milano, Milan, Italy"],"raw_orcid":"https://orcid.org/0009-0009-3025-8620","affiliations":[{"raw_affiliation_string":"Politecnico di Milano, Milan, Italy","institution_ids":["https://openalex.org/I93860229"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5092631473","display_name":"Benjamin Ramhorst","orcid":"https://orcid.org/0000-0002-0026-1281"},"institutions":[{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Benjamin Ramhorst","raw_affiliation_strings":["ETH Zurich, Zurich, Switzerland"],"raw_orcid":"https://orcid.org/0000-0002-0026-1281","affiliations":[{"raw_affiliation_string":"ETH Zurich, Zurich, Switzerland","institution_ids":["https://openalex.org/I35440088"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5035659790","display_name":"Maximilian Jakob Heer","orcid":"https://orcid.org/0009-0000-9231-0089"},"institutions":[{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Maximilian Jakob Heer","raw_affiliation_strings":["ETH Zurich, Zurich, Switzerland"],"raw_orcid":"https://orcid.org/0009-0000-9231-0089","affiliations":[{"raw_affiliation_string":"ETH Zurich, Zurich, Switzerland","institution_ids":["https://openalex.org/I35440088"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5062570405","display_name":"Lucian Petric\u0103","orcid":"https://orcid.org/0000-0002-3480-0570"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lucian Petrica","raw_affiliation_strings":["AMD Research and Advanced Development, Dublin, Ireland"],"raw_orcid":"https://orcid.org/0000-0002-3480-0570","affiliations":[{"raw_affiliation_string":"AMD Research and Advanced Development, Dublin, Ireland","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5068982084","display_name":"Dario Korolija","orcid":"https://orcid.org/0009-0002-3984-3948"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dario Korolija","raw_affiliation_strings":["AMD Research and Advanced Development, Dublin, Ireland"],"raw_orcid":"https://orcid.org/0009-0002-3984-3948","affiliations":[{"raw_affiliation_string":"AMD Research and Advanced Development, Dublin, Ireland","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5010543929","display_name":"Marco D. Santambrogio","orcid":"https://orcid.org/0000-0002-9883-9693"},"institutions":[{"id":"https://openalex.org/I93860229","display_name":"Politecnico di Milano","ror":"https://ror.org/01nffqt88","country_code":"IT","type":"education","lineage":["https://openalex.org/I93860229"]}],"countries":["IT"],"is_corresponding":false,"raw_author_name":"Marco D. Santambrogio","raw_affiliation_strings":["Politecnico di Milano, Milan, Italy"],"raw_orcid":"https://orcid.org/0000-0002-9883-9693","affiliations":[{"raw_affiliation_string":"Politecnico di Milano, Milan, Italy","institution_ids":["https://openalex.org/I93860229"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5031247318","display_name":"Davide Conficconi","orcid":"https://orcid.org/0000-0002-5834-0812"},"institutions":[{"id":"https://openalex.org/I93860229","display_name":"Politecnico di Milano","ror":"https://ror.org/01nffqt88","country_code":"IT","type":"education","lineage":["https://openalex.org/I93860229"]}],"countries":["IT"],"is_corresponding":false,"raw_author_name":"Davide Conficconi","raw_affiliation_strings":["Politecnico di Milano, Milan, Italy"],"raw_orcid":"https://orcid.org/0000-0002-5834-0812","affiliations":[{"raw_affiliation_string":"Politecnico di Milano, Milan, Italy","institution_ids":["https://openalex.org/I93860229"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103144919","display_name":"Gustavo Alonso","orcid":"https://orcid.org/0000-0002-4396-6695"},"institutions":[{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Gustavo Alonso","raw_affiliation_strings":["ETH Zurich, Zurich, Switzerland"],"raw_orcid":"https://orcid.org/0000-0002-4396-6695","affiliations":[{"raw_affiliation_string":"ETH Zurich, Zurich, Switzerland","institution_ids":["https://openalex.org/I35440088"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5134500116","display_name":"Ken O'Brien","orcid":"https://orcid.org/0000-0002-5644-2110"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ken O'Brien","raw_affiliation_strings":["AMD Research and Advanced Development, Dublin, Ireland"],"raw_orcid":"https://orcid.org/0000-0002-5644-2110","affiliations":[{"raw_affiliation_string":"AMD Research and Advanced Development, Dublin, Ireland","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":18.0163,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.98013544,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":94,"max":97},"biblio":{"volume":null,"issue":null,"first_page":"1829","last_page":"1846"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.30559998750686646,"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":0.30559998750686646,"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/T10101","display_name":"Cloud Computing and Resource Management","score":0.1356000006198883,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T10742","display_name":"Peer-to-Peer Network Technologies","score":0.06800000369548798,"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/field-programmable-gate-array","display_name":"Field-programmable gate array","score":0.44850000739097595},{"id":"https://openalex.org/keywords/architecture","display_name":"Architecture","score":0.39079999923706055},{"id":"https://openalex.org/keywords/reconfigurable-computing","display_name":"Reconfigurable computing","score":0.29679998755455017},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.28999999165534973},{"id":"https://openalex.org/keywords/systems-architecture","display_name":"Systems architecture","score":0.2786000072956085}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6229000091552734},{"id":"https://openalex.org/C42935608","wikidata":"https://www.wikidata.org/wiki/Q190411","display_name":"Field-programmable gate array","level":2,"score":0.44850000739097595},{"id":"https://openalex.org/C123657996","wikidata":"https://www.wikidata.org/wiki/Q12271","display_name":"Architecture","level":2,"score":0.39079999923706055},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.37779998779296875},{"id":"https://openalex.org/C142962650","wikidata":"https://www.wikidata.org/wiki/Q240838","display_name":"Reconfigurable computing","level":3,"score":0.29679998755455017},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.2953000068664551},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.28999999165534973},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.28600001335144043},{"id":"https://openalex.org/C98025372","wikidata":"https://www.wikidata.org/wiki/Q477538","display_name":"Systems architecture","level":3,"score":0.2786000072956085},{"id":"https://openalex.org/C459310","wikidata":"https://www.wikidata.org/wiki/Q117801","display_name":"Computational science","level":1,"score":0.26930001378059387},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.26159998774528503},{"id":"https://openalex.org/C77618280","wikidata":"https://www.wikidata.org/wiki/Q1155772","display_name":"Scheme (mathematics)","level":2,"score":0.25450000166893005}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1145/3767295.3803620","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3767295.3803620","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 21st European Conference on Computer Systems","raw_type":"proceedings-article"},{"id":"pmh:oai:re.public.polimi.it:11311/1319597","is_oa":false,"landing_page_url":"https://hdl.handle.net/11311/1319597","pdf_url":null,"source":{"id":"https://openalex.org/S4306400312","display_name":"Virtual Community of Pathological Anatomy (University of Castilla La Mancha)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I79189158","host_organization_name":"University of Castilla-La Mancha","host_organization_lineage":["https://openalex.org/I79189158"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"info:eu-repo/semantics/conferenceObject"}],"best_oa_location":{"id":"doi:10.1145/3767295.3803620","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3767295.3803620","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 21st European Conference on Computer Systems","raw_type":"proceedings-article"},"sustainable_development_goals":[{"display_name":"Industry, innovation and infrastructure","score":0.4412960112094879,"id":"https://metadata.un.org/sdg/9"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":24,"referenced_works":["https://openalex.org/W2002227246","https://openalex.org/W2083927153","https://openalex.org/W2133287637","https://openalex.org/W2885978796","https://openalex.org/W2926767350","https://openalex.org/W2941091233","https://openalex.org/W2997709574","https://openalex.org/W3215120999","https://openalex.org/W4224952058","https://openalex.org/W4225647032","https://openalex.org/W4225759418","https://openalex.org/W4286203032","https://openalex.org/W4310551570","https://openalex.org/W4312783897","https://openalex.org/W4377942948","https://openalex.org/W4386569012","https://openalex.org/W4387546476","https://openalex.org/W4392699793","https://openalex.org/W4399677840","https://openalex.org/W4401879382","https://openalex.org/W4405375424","https://openalex.org/W4412380935","https://openalex.org/W6913065645","https://openalex.org/W6958896998"],"related_works":[],"abstract_inverted_index":{"Modern":[0],"datacenters":[1],"integrate":[2],"heterogeneous":[3],"accelerators,":[4],"such":[5,188],"as":[6,189],"GPUs":[7,18,120],"and":[8,37,61,88,121,136,166,191,193,200,219],"FPGAs,":[9],"to":[10,57,78,93,126,179,263],"speed":[11],"up":[12,262],"different":[13,149,207],"stages":[14],"of":[15],"compute-intensive":[16,237],"pipelines.":[17],"are":[19,73],"best":[20],"suited":[21],"for":[22,111,154,160,235],"massively":[23],"parallel":[24],"workloads":[25],"(e.g.,":[26],"deep":[27],"learning),":[28],"while":[29],"FPGAs":[30,190],"excel":[31],"at":[32],"task-level":[33],"parallelism,":[34],"stream-oriented":[35],"processing,":[36],"in-network":[38],"acceleration.":[39],"Since":[40],"these":[41],"architectures":[42],"must":[43],"exchange":[44],"data":[45,228,269],"efficiently,":[46],"literature":[47],"introduced":[48],"Peer-To-Peer":[49],"(P2P)":[50],"communication":[51,83,114,185,213,240],"across":[52,84,148,206],"PCI":[53],"Express":[54],"(PCIe)":[55],"devices,":[56],"reduce":[58],"CPU-driven":[59],"orchestration":[60],"avoid":[62],"intermediate,":[63],"redundant":[64],"buffer":[65],"copies":[66],"that":[67,132,182],"degrade":[68],"performance.":[69],"However,":[70],"current":[71],"solutions":[72],"either":[74],"closed-source":[75],"or":[76],"tied":[77],"proprietary":[79],"frameworks,":[80],"limiting":[81],"P2P":[82,95,112],"most":[85],"PCIe-based":[86],"devices":[87],"requiring":[89],"significant":[90],"technical":[91],"effort":[92],"enable":[94],"capabilities":[96],"on":[97,118],"supported":[98],"hardware.":[99],"For":[100],"this":[101,139],"reason,":[102],"we":[103,141,210,231,255],"propose":[104],"RoPeerTo,":[105],"a":[106,128,143,172,176,216,220,236,243,247,264],"fully":[107],"open-source,":[108],"datacenter-scale":[109],"architecture":[110],"DMA":[113],"over":[115],"PCIe,":[116],"validated":[117],"both":[119],"FPGAs.":[122],"The":[123,169],"goal":[124],"is":[125,171,241],"provide":[127],"general,":[129],"open":[130],"alternative":[131],"ensures":[133],"flexibility,":[134],"efficiency,":[135],"usability.":[137],"To":[138],"end,":[140],"design":[142],"complete":[144],"HW/SW":[145],"stack":[146],"operating":[147],"layers,":[150],"supporting":[151],"standard":[152],"protocols":[153],"DMA-based":[155],"memory":[156,163],"sharing,":[157],"advanced":[158],"tools":[159],"device":[161],"virtualization,":[162],"address":[164],"translation,":[165],"access":[167],"protection.":[168],"result":[170],"unified":[173],"framework":[174],"exposing":[175],"high-level":[177],"API":[178],"end":[180],"users,":[181],"enables":[183],"direct":[184],"between":[186],"accelerators":[187],"GPUs,":[192],"abstracts":[194],"away":[195],"the":[196,204,212,233],"underlying":[197],"hardware":[198],"setup":[199],"management.":[201],"We":[202],"validate":[203],"system":[205,234],"scenarios.":[208],"First,":[209],"isolate":[211],"layer,":[214],"observing":[215],"5.61\u00d7":[217],"speedup":[218,249,266],"37.99%":[221],"reduction":[222],"in":[223,267],"GPU":[224],"power":[225],"consumption":[226],"during":[227],"transfer.":[229],"Next,":[230],"leverage":[232],"workload":[238],"where":[239],"only":[242],"partial":[244],"bottleneck,":[245],"achieving":[246],"6.77%":[248],"without":[250],"any":[251],"compute-side":[252],"modifications.":[253],"Finally,":[254],"evaluate":[256],"communication-heavy":[257],"distributed":[258],"computing":[259],"workloads,":[260],"demonstrating":[261],"21.79\u00d7":[265],"network-bound":[268],"scattering.":[270]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-04-25T00:00:00"}
