On Optimizing the Communication of Model Parallelism Article Swipe
YOU?
·
· 2022
· Open Access
·
· DOI: https://doi.org/10.48550/arxiv.2211.05322
We study a novel and important communication pattern in large-scale model-parallel deep learning (DL), which we call cross-mesh resharding. This pattern emerges when the two paradigms of model parallelism - intra-operator and inter-operator parallelism - are combined to support large models on large clusters. In cross-mesh resharding, a sharded tensor needs to be sent from a source device mesh to a destination device mesh, on which the tensor may be distributed with the same or different layouts. We formalize this as a many-to-many multicast communication problem, and show that existing approaches either are sub-optimal or do not generalize to different network topologies or tensor layouts, which result from different model architectures and parallelism strategies. We then propose two contributions to address cross-mesh resharding: an efficient broadcast-based communication system, and an "overlapping-friendly" pipeline schedule. On microbenchmarks, our overall system outperforms existing ones by up to 10x across various tensor and mesh layouts. On end-to-end training of two large models, GPT-3 and U-Transformer, we improve throughput by 10% and 50%, respectively.
Related Topics
- Type
- preprint
- Language
- en
- Landing Page
- http://arxiv.org/abs/2211.05322
- https://arxiv.org/pdf/2211.05322
- OA Status
- green
- Cited By
- 3
- Related Works
- 10
- OpenAlex ID
- https://openalex.org/W4308828369
Raw OpenAlex JSON
- OpenAlex ID
-
https://openalex.org/W4308828369Canonical identifier for this work in OpenAlex
- DOI
-
https://doi.org/10.48550/arxiv.2211.05322Digital Object Identifier
- Title
-
On Optimizing the Communication of Model ParallelismWork title
- Type
-
preprintOpenAlex work type
- Language
-
enPrimary language
- Publication year
-
2022Year of publication
- Publication date
-
2022-11-10Full publication date if available
- Authors
-
Yonghao Zhuang, Hexu Zhao, Lianmin Zheng, Zhuohan Li, Eric P. Xing, Qirong Ho, Joseph E. Gonzalez, Ion Stoica, Hao ZhangList of authors in order
- Landing page
-
https://arxiv.org/abs/2211.05322Publisher landing page
- PDF URL
-
https://arxiv.org/pdf/2211.05322Direct link to full text PDF
- Open access
-
YesWhether a free full text is available
- OA status
-
greenOpen access status per OpenAlex
- OA URL
-
https://arxiv.org/pdf/2211.05322Direct OA link when available
- Concepts
-
Computer science, Multicast, Parallelism (grammar), Parallel computing, Distributed computing, Network topology, Models of communication, Pipeline (software), Operator (biology), Schedule, Theoretical computer science, Computer network, Programming language, Operating system, Gene, Repressor, Chemistry, Biochemistry, Communication, Sociology, Transcription factorTop concepts (fields/topics) attached by OpenAlex
- Cited by
-
3Total citation count in OpenAlex
- Citations by year (recent)
-
2025: 1, 2024: 2Per-year citation counts (last 5 years)
- Related works (count)
-
10Other works algorithmically related by OpenAlex
Full payload
| id | https://openalex.org/W4308828369 |
|---|---|
| doi | https://doi.org/10.48550/arxiv.2211.05322 |
| ids.doi | https://doi.org/10.48550/arxiv.2211.05322 |
| ids.openalex | https://openalex.org/W4308828369 |
| fwci | |
| type | preprint |
| title | On Optimizing the Communication of Model Parallelism |
| biblio.issue | |
| biblio.volume | |
| biblio.last_page | |
| biblio.first_page | |
| topics[0].id | https://openalex.org/T10036 |
| topics[0].field.id | https://openalex.org/fields/17 |
| topics[0].field.display_name | Computer Science |
| topics[0].score | 0.9958000183105469 |
| topics[0].domain.id | https://openalex.org/domains/3 |
| topics[0].domain.display_name | Physical Sciences |
| topics[0].subfield.id | https://openalex.org/subfields/1707 |
| topics[0].subfield.display_name | Computer Vision and Pattern Recognition |
| topics[0].display_name | Advanced Neural Network Applications |
| topics[1].id | https://openalex.org/T12808 |
| topics[1].field.id | https://openalex.org/fields/22 |
| topics[1].field.display_name | Engineering |
| topics[1].score | 0.994700014591217 |
| topics[1].domain.id | https://openalex.org/domains/3 |
| topics[1].domain.display_name | Physical Sciences |
| topics[1].subfield.id | https://openalex.org/subfields/2208 |
| topics[1].subfield.display_name | Electrical and Electronic Engineering |
| topics[1].display_name | Ferroelectric and Negative Capacitance Devices |
| topics[2].id | https://openalex.org/T10054 |
| topics[2].field.id | https://openalex.org/fields/17 |
| topics[2].field.display_name | Computer Science |
| topics[2].score | 0.993399977684021 |
| topics[2].domain.id | https://openalex.org/domains/3 |
| topics[2].domain.display_name | Physical Sciences |
| topics[2].subfield.id | https://openalex.org/subfields/1708 |
| topics[2].subfield.display_name | Hardware and Architecture |
| topics[2].display_name | Parallel Computing and Optimization Techniques |
| is_xpac | False |
| apc_list | |
| apc_paid | |
| concepts[0].id | https://openalex.org/C41008148 |
| concepts[0].level | 0 |
| concepts[0].score | 0.8460873365402222 |
| concepts[0].wikidata | https://www.wikidata.org/wiki/Q21198 |
| concepts[0].display_name | Computer science |
| concepts[1].id | https://openalex.org/C32295351 |
| concepts[1].level | 2 |
| concepts[1].score | 0.7035799026489258 |
| concepts[1].wikidata | https://www.wikidata.org/wiki/Q899288 |
| concepts[1].display_name | Multicast |
| concepts[2].id | https://openalex.org/C2781172179 |
| concepts[2].level | 2 |
| concepts[2].score | 0.683171808719635 |
| concepts[2].wikidata | https://www.wikidata.org/wiki/Q853109 |
| concepts[2].display_name | Parallelism (grammar) |
| concepts[3].id | https://openalex.org/C173608175 |
| concepts[3].level | 1 |
| concepts[3].score | 0.6204637289047241 |
| concepts[3].wikidata | https://www.wikidata.org/wiki/Q232661 |
| concepts[3].display_name | Parallel computing |
| concepts[4].id | https://openalex.org/C120314980 |
| concepts[4].level | 1 |
| concepts[4].score | 0.4838391840457916 |
| concepts[4].wikidata | https://www.wikidata.org/wiki/Q180634 |
| concepts[4].display_name | Distributed computing |
| concepts[5].id | https://openalex.org/C199845137 |
| concepts[5].level | 2 |
| concepts[5].score | 0.4813588261604309 |
| concepts[5].wikidata | https://www.wikidata.org/wiki/Q145490 |
| concepts[5].display_name | Network topology |
| concepts[6].id | https://openalex.org/C158156997 |
| concepts[6].level | 2 |
| concepts[6].score | 0.4491894841194153 |
| concepts[6].wikidata | https://www.wikidata.org/wiki/Q1416645 |
| concepts[6].display_name | Models of communication |
| concepts[7].id | https://openalex.org/C43521106 |
| concepts[7].level | 2 |
| concepts[7].score | 0.4304533898830414 |
| concepts[7].wikidata | https://www.wikidata.org/wiki/Q2165493 |
| concepts[7].display_name | Pipeline (software) |
| concepts[8].id | https://openalex.org/C17020691 |
| concepts[8].level | 5 |
| concepts[8].score | 0.41290047764778137 |
| concepts[8].wikidata | https://www.wikidata.org/wiki/Q139677 |
| concepts[8].display_name | Operator (biology) |
| concepts[9].id | https://openalex.org/C68387754 |
| concepts[9].level | 2 |
| concepts[9].score | 0.4122838079929352 |
| concepts[9].wikidata | https://www.wikidata.org/wiki/Q7271585 |
| concepts[9].display_name | Schedule |
| concepts[10].id | https://openalex.org/C80444323 |
| concepts[10].level | 1 |
| concepts[10].score | 0.3747384250164032 |
| concepts[10].wikidata | https://www.wikidata.org/wiki/Q2878974 |
| concepts[10].display_name | Theoretical computer science |
| concepts[11].id | https://openalex.org/C31258907 |
| concepts[11].level | 1 |
| concepts[11].score | 0.18045195937156677 |
| concepts[11].wikidata | https://www.wikidata.org/wiki/Q1301371 |
| concepts[11].display_name | Computer network |
| concepts[12].id | https://openalex.org/C199360897 |
| concepts[12].level | 1 |
| concepts[12].score | 0.10308873653411865 |
| concepts[12].wikidata | https://www.wikidata.org/wiki/Q9143 |
| concepts[12].display_name | Programming language |
| concepts[13].id | https://openalex.org/C111919701 |
| concepts[13].level | 1 |
| concepts[13].score | 0.07229351997375488 |
| concepts[13].wikidata | https://www.wikidata.org/wiki/Q9135 |
| concepts[13].display_name | Operating system |
| concepts[14].id | https://openalex.org/C104317684 |
| concepts[14].level | 2 |
| concepts[14].score | 0.0 |
| concepts[14].wikidata | https://www.wikidata.org/wiki/Q7187 |
| concepts[14].display_name | Gene |
| concepts[15].id | https://openalex.org/C158448853 |
| concepts[15].level | 4 |
| concepts[15].score | 0.0 |
| concepts[15].wikidata | https://www.wikidata.org/wiki/Q425218 |
| concepts[15].display_name | Repressor |
| concepts[16].id | https://openalex.org/C185592680 |
| concepts[16].level | 0 |
| concepts[16].score | 0.0 |
| concepts[16].wikidata | https://www.wikidata.org/wiki/Q2329 |
| concepts[16].display_name | Chemistry |
| concepts[17].id | https://openalex.org/C55493867 |
| concepts[17].level | 1 |
| concepts[17].score | 0.0 |
| concepts[17].wikidata | https://www.wikidata.org/wiki/Q7094 |
| concepts[17].display_name | Biochemistry |
| concepts[18].id | https://openalex.org/C46312422 |
| concepts[18].level | 1 |
| concepts[18].score | 0.0 |
| concepts[18].wikidata | https://www.wikidata.org/wiki/Q11024 |
| concepts[18].display_name | Communication |
| concepts[19].id | https://openalex.org/C144024400 |
| concepts[19].level | 0 |
| concepts[19].score | 0.0 |
| concepts[19].wikidata | https://www.wikidata.org/wiki/Q21201 |
| concepts[19].display_name | Sociology |
| concepts[20].id | https://openalex.org/C86339819 |
| concepts[20].level | 3 |
| concepts[20].score | 0.0 |
| concepts[20].wikidata | https://www.wikidata.org/wiki/Q407384 |
| concepts[20].display_name | Transcription factor |
| keywords[0].id | https://openalex.org/keywords/computer-science |
| keywords[0].score | 0.8460873365402222 |
| keywords[0].display_name | Computer science |
| keywords[1].id | https://openalex.org/keywords/multicast |
| keywords[1].score | 0.7035799026489258 |
| keywords[1].display_name | Multicast |
| keywords[2].id | https://openalex.org/keywords/parallelism |
| keywords[2].score | 0.683171808719635 |
| keywords[2].display_name | Parallelism (grammar) |
| keywords[3].id | https://openalex.org/keywords/parallel-computing |
| keywords[3].score | 0.6204637289047241 |
| keywords[3].display_name | Parallel computing |
| keywords[4].id | https://openalex.org/keywords/distributed-computing |
| keywords[4].score | 0.4838391840457916 |
| keywords[4].display_name | Distributed computing |
| keywords[5].id | https://openalex.org/keywords/network-topology |
| keywords[5].score | 0.4813588261604309 |
| keywords[5].display_name | Network topology |
| keywords[6].id | https://openalex.org/keywords/models-of-communication |
| keywords[6].score | 0.4491894841194153 |
| keywords[6].display_name | Models of communication |
| keywords[7].id | https://openalex.org/keywords/pipeline |
| keywords[7].score | 0.4304533898830414 |
| keywords[7].display_name | Pipeline (software) |
| keywords[8].id | https://openalex.org/keywords/operator |
| keywords[8].score | 0.41290047764778137 |
| keywords[8].display_name | Operator (biology) |
| keywords[9].id | https://openalex.org/keywords/schedule |
| keywords[9].score | 0.4122838079929352 |
| keywords[9].display_name | Schedule |
| keywords[10].id | https://openalex.org/keywords/theoretical-computer-science |
| keywords[10].score | 0.3747384250164032 |
| keywords[10].display_name | Theoretical computer science |
| keywords[11].id | https://openalex.org/keywords/computer-network |
| keywords[11].score | 0.18045195937156677 |
| keywords[11].display_name | Computer network |
| keywords[12].id | https://openalex.org/keywords/programming-language |
| keywords[12].score | 0.10308873653411865 |
| keywords[12].display_name | Programming language |
| keywords[13].id | https://openalex.org/keywords/operating-system |
| keywords[13].score | 0.07229351997375488 |
| keywords[13].display_name | Operating system |
| language | en |
| locations[0].id | pmh:oai:arXiv.org:2211.05322 |
| locations[0].is_oa | True |
| locations[0].source.id | https://openalex.org/S4306400194 |
| locations[0].source.issn | |
| locations[0].source.type | repository |
| locations[0].source.is_oa | True |
| locations[0].source.issn_l | |
| locations[0].source.is_core | False |
| locations[0].source.is_in_doaj | False |
| locations[0].source.display_name | arXiv (Cornell University) |
| locations[0].source.host_organization | https://openalex.org/I205783295 |
| locations[0].source.host_organization_name | Cornell University |
| locations[0].source.host_organization_lineage | https://openalex.org/I205783295 |
| locations[0].license | |
| locations[0].pdf_url | https://arxiv.org/pdf/2211.05322 |
| locations[0].version | submittedVersion |
| locations[0].raw_type | text |
| locations[0].license_id | |
| locations[0].is_accepted | False |
| locations[0].is_published | False |
| locations[0].raw_source_name | |
| locations[0].landing_page_url | http://arxiv.org/abs/2211.05322 |
| locations[1].id | doi:10.48550/arxiv.2211.05322 |
| locations[1].is_oa | True |
| locations[1].source.id | https://openalex.org/S4306400194 |
| locations[1].source.issn | |
| locations[1].source.type | repository |
| locations[1].source.is_oa | True |
| locations[1].source.issn_l | |
| locations[1].source.is_core | False |
| locations[1].source.is_in_doaj | False |
| locations[1].source.display_name | arXiv (Cornell University) |
| locations[1].source.host_organization | https://openalex.org/I205783295 |
| locations[1].source.host_organization_name | Cornell University |
| locations[1].source.host_organization_lineage | https://openalex.org/I205783295 |
| locations[1].license | cc-by |
| locations[1].pdf_url | |
| locations[1].version | |
| locations[1].raw_type | article |
| locations[1].license_id | https://openalex.org/licenses/cc-by |
| locations[1].is_accepted | False |
| locations[1].is_published | |
| locations[1].raw_source_name | |
| locations[1].landing_page_url | https://doi.org/10.48550/arxiv.2211.05322 |
| indexed_in | arxiv, datacite |
| authorships[0].author.id | https://openalex.org/A5076407338 |
| authorships[0].author.orcid | |
| authorships[0].author.display_name | Yonghao Zhuang |
| authorships[0].author_position | first |
| authorships[0].raw_author_name | Zhuang, Yonghao |
| authorships[0].is_corresponding | False |
| authorships[1].author.id | https://openalex.org/A5030301481 |
| authorships[1].author.orcid | |
| authorships[1].author.display_name | Hexu Zhao |
| authorships[1].author_position | middle |
| authorships[1].raw_author_name | Zhao, Hexu |
| authorships[1].is_corresponding | False |
| authorships[2].author.id | https://openalex.org/A5061339425 |
| authorships[2].author.orcid | https://orcid.org/0000-0001-5812-731X |
| authorships[2].author.display_name | Lianmin Zheng |
| authorships[2].author_position | middle |
| authorships[2].raw_author_name | Zheng, Lianmin |
| authorships[2].is_corresponding | False |
| authorships[3].author.id | https://openalex.org/A5009362721 |
| authorships[3].author.orcid | https://orcid.org/0000-0001-5372-9450 |
| authorships[3].author.display_name | Zhuohan Li |
| authorships[3].author_position | middle |
| authorships[3].raw_author_name | Li, Zhuohan |
| authorships[3].is_corresponding | False |
| authorships[4].author.id | https://openalex.org/A5009547049 |
| authorships[4].author.orcid | https://orcid.org/0009-0005-9158-4201 |
| authorships[4].author.display_name | Eric P. Xing |
| authorships[4].author_position | middle |
| authorships[4].raw_author_name | Xing, Eric P. |
| authorships[4].is_corresponding | False |
| authorships[5].author.id | https://openalex.org/A5012361506 |
| authorships[5].author.orcid | |
| authorships[5].author.display_name | Qirong Ho |
| authorships[5].author_position | middle |
| authorships[5].raw_author_name | Ho, Qirong |
| authorships[5].is_corresponding | False |
| authorships[6].author.id | https://openalex.org/A5072427753 |
| authorships[6].author.orcid | https://orcid.org/0000-0003-2921-956X |
| authorships[6].author.display_name | Joseph E. Gonzalez |
| authorships[6].author_position | middle |
| authorships[6].raw_author_name | Gonzalez, Joseph E. |
| authorships[6].is_corresponding | False |
| authorships[7].author.id | https://openalex.org/A5041920173 |
| authorships[7].author.orcid | https://orcid.org/0000-0002-5373-0088 |
| authorships[7].author.display_name | Ion Stoica |
| authorships[7].author_position | middle |
| authorships[7].raw_author_name | Stoica, Ion |
| authorships[7].is_corresponding | False |
| authorships[8].author.id | https://openalex.org/A5089432787 |
| authorships[8].author.orcid | https://orcid.org/0000-0003-1923-589X |
| authorships[8].author.display_name | Hao Zhang |
| authorships[8].author_position | last |
| authorships[8].raw_author_name | Zhang, Hao |
| authorships[8].is_corresponding | False |
| has_content.pdf | False |
| has_content.grobid_xml | False |
| is_paratext | False |
| open_access.is_oa | True |
| open_access.oa_url | https://arxiv.org/pdf/2211.05322 |
| open_access.oa_status | green |
| open_access.any_repository_has_fulltext | False |
| created_date | 2025-10-10T00:00:00 |
| display_name | On Optimizing the Communication of Model Parallelism |
| has_fulltext | False |
| is_retracted | False |
| updated_date | 2025-11-06T06:51:31.235846 |
| primary_topic.id | https://openalex.org/T10036 |
| primary_topic.field.id | https://openalex.org/fields/17 |
| primary_topic.field.display_name | Computer Science |
| primary_topic.score | 0.9958000183105469 |
| primary_topic.domain.id | https://openalex.org/domains/3 |
| primary_topic.domain.display_name | Physical Sciences |
| primary_topic.subfield.id | https://openalex.org/subfields/1707 |
| primary_topic.subfield.display_name | Computer Vision and Pattern Recognition |
| primary_topic.display_name | Advanced Neural Network Applications |
| related_works | https://openalex.org/W2033488476, https://openalex.org/W2131092765, https://openalex.org/W1589992863, https://openalex.org/W2277662012, https://openalex.org/W1617457131, https://openalex.org/W4377015086, https://openalex.org/W2369090769, https://openalex.org/W2997509936, https://openalex.org/W136477243, https://openalex.org/W2033862586 |
| cited_by_count | 3 |
| counts_by_year[0].year | 2025 |
| counts_by_year[0].cited_by_count | 1 |
| counts_by_year[1].year | 2024 |
| counts_by_year[1].cited_by_count | 2 |
| locations_count | 2 |
| best_oa_location.id | pmh:oai:arXiv.org:2211.05322 |
| best_oa_location.is_oa | True |
| best_oa_location.source.id | https://openalex.org/S4306400194 |
| best_oa_location.source.issn | |
| best_oa_location.source.type | repository |
| best_oa_location.source.is_oa | True |
| best_oa_location.source.issn_l | |
| best_oa_location.source.is_core | False |
| best_oa_location.source.is_in_doaj | False |
| best_oa_location.source.display_name | arXiv (Cornell University) |
| best_oa_location.source.host_organization | https://openalex.org/I205783295 |
| best_oa_location.source.host_organization_name | Cornell University |
| best_oa_location.source.host_organization_lineage | https://openalex.org/I205783295 |
| best_oa_location.license | |
| best_oa_location.pdf_url | https://arxiv.org/pdf/2211.05322 |
| best_oa_location.version | submittedVersion |
| best_oa_location.raw_type | text |
| best_oa_location.license_id | |
| best_oa_location.is_accepted | False |
| best_oa_location.is_published | False |
| best_oa_location.raw_source_name | |
| best_oa_location.landing_page_url | http://arxiv.org/abs/2211.05322 |
| primary_location.id | pmh:oai:arXiv.org:2211.05322 |
| primary_location.is_oa | True |
| primary_location.source.id | https://openalex.org/S4306400194 |
| primary_location.source.issn | |
| primary_location.source.type | repository |
| primary_location.source.is_oa | True |
| primary_location.source.issn_l | |
| primary_location.source.is_core | False |
| primary_location.source.is_in_doaj | False |
| primary_location.source.display_name | arXiv (Cornell University) |
| primary_location.source.host_organization | https://openalex.org/I205783295 |
| primary_location.source.host_organization_name | Cornell University |
| primary_location.source.host_organization_lineage | https://openalex.org/I205783295 |
| primary_location.license | |
| primary_location.pdf_url | https://arxiv.org/pdf/2211.05322 |
| primary_location.version | submittedVersion |
| primary_location.raw_type | text |
| primary_location.license_id | |
| primary_location.is_accepted | False |
| primary_location.is_published | False |
| primary_location.raw_source_name | |
| primary_location.landing_page_url | http://arxiv.org/abs/2211.05322 |
| publication_date | 2022-11-10 |
| publication_year | 2022 |
| referenced_works_count | 0 |
| abstract_inverted_index.- | 29, 34 |
| abstract_inverted_index.a | 2, 47, 55, 60, 81 |
| abstract_inverted_index.In | 44 |
| abstract_inverted_index.On | 133, 151 |
| abstract_inverted_index.We | 0, 77, 114 |
| abstract_inverted_index.an | 123, 129 |
| abstract_inverted_index.as | 80 |
| abstract_inverted_index.be | 52, 69 |
| abstract_inverted_index.by | 141, 164 |
| abstract_inverted_index.do | 95 |
| abstract_inverted_index.in | 8 |
| abstract_inverted_index.of | 26, 154 |
| abstract_inverted_index.on | 41, 64 |
| abstract_inverted_index.or | 74, 94, 102 |
| abstract_inverted_index.to | 37, 51, 59, 98, 119, 143 |
| abstract_inverted_index.up | 142 |
| abstract_inverted_index.we | 15, 161 |
| abstract_inverted_index.10% | 165 |
| abstract_inverted_index.10x | 144 |
| abstract_inverted_index.and | 4, 31, 86, 111, 128, 148, 159, 166 |
| abstract_inverted_index.are | 35, 92 |
| abstract_inverted_index.may | 68 |
| abstract_inverted_index.not | 96 |
| abstract_inverted_index.our | 135 |
| abstract_inverted_index.the | 23, 66, 72 |
| abstract_inverted_index.two | 24, 117, 155 |
| abstract_inverted_index.50%, | 167 |
| abstract_inverted_index.This | 19 |
| abstract_inverted_index.call | 16 |
| abstract_inverted_index.deep | 11 |
| abstract_inverted_index.from | 54, 107 |
| abstract_inverted_index.mesh | 58, 149 |
| abstract_inverted_index.ones | 140 |
| abstract_inverted_index.same | 73 |
| abstract_inverted_index.sent | 53 |
| abstract_inverted_index.show | 87 |
| abstract_inverted_index.that | 88 |
| abstract_inverted_index.then | 115 |
| abstract_inverted_index.this | 79 |
| abstract_inverted_index.when | 22 |
| abstract_inverted_index.with | 71 |
| abstract_inverted_index.(DL), | 13 |
| abstract_inverted_index.GPT-3 | 158 |
| abstract_inverted_index.large | 39, 42, 156 |
| abstract_inverted_index.mesh, | 63 |
| abstract_inverted_index.model | 27, 109 |
| abstract_inverted_index.needs | 50 |
| abstract_inverted_index.novel | 3 |
| abstract_inverted_index.study | 1 |
| abstract_inverted_index.which | 14, 65, 105 |
| abstract_inverted_index.across | 145 |
| abstract_inverted_index.device | 57, 62 |
| abstract_inverted_index.either | 91 |
| abstract_inverted_index.models | 40 |
| abstract_inverted_index.result | 106 |
| abstract_inverted_index.source | 56 |
| abstract_inverted_index.system | 137 |
| abstract_inverted_index.tensor | 49, 67, 103, 147 |
| abstract_inverted_index.address | 120 |
| abstract_inverted_index.emerges | 21 |
| abstract_inverted_index.improve | 162 |
| abstract_inverted_index.models, | 157 |
| abstract_inverted_index.network | 100 |
| abstract_inverted_index.overall | 136 |
| abstract_inverted_index.pattern | 7, 20 |
| abstract_inverted_index.propose | 116 |
| abstract_inverted_index.sharded | 48 |
| abstract_inverted_index.support | 38 |
| abstract_inverted_index.system, | 127 |
| abstract_inverted_index.various | 146 |
| abstract_inverted_index.combined | 36 |
| abstract_inverted_index.existing | 89, 139 |
| abstract_inverted_index.layouts, | 104 |
| abstract_inverted_index.layouts. | 76, 150 |
| abstract_inverted_index.learning | 12 |
| abstract_inverted_index.pipeline | 131 |
| abstract_inverted_index.problem, | 85 |
| abstract_inverted_index.training | 153 |
| abstract_inverted_index.clusters. | 43 |
| abstract_inverted_index.different | 75, 99, 108 |
| abstract_inverted_index.efficient | 124 |
| abstract_inverted_index.formalize | 78 |
| abstract_inverted_index.important | 5 |
| abstract_inverted_index.multicast | 83 |
| abstract_inverted_index.paradigms | 25 |
| abstract_inverted_index.schedule. | 132 |
| abstract_inverted_index.approaches | 90 |
| abstract_inverted_index.cross-mesh | 17, 45, 121 |
| abstract_inverted_index.end-to-end | 152 |
| abstract_inverted_index.generalize | 97 |
| abstract_inverted_index.throughput | 163 |
| abstract_inverted_index.topologies | 101 |
| abstract_inverted_index.destination | 61 |
| abstract_inverted_index.distributed | 70 |
| abstract_inverted_index.large-scale | 9 |
| abstract_inverted_index.outperforms | 138 |
| abstract_inverted_index.parallelism | 28, 33, 112 |
| abstract_inverted_index.resharding, | 46 |
| abstract_inverted_index.resharding. | 18 |
| abstract_inverted_index.resharding: | 122 |
| abstract_inverted_index.strategies. | 113 |
| abstract_inverted_index.sub-optimal | 93 |
| abstract_inverted_index.many-to-many | 82 |
| abstract_inverted_index.architectures | 110 |
| abstract_inverted_index.communication | 6, 84, 126 |
| abstract_inverted_index.contributions | 118 |
| abstract_inverted_index.respectively. | 168 |
| abstract_inverted_index.U-Transformer, | 160 |
| abstract_inverted_index.inter-operator | 32 |
| abstract_inverted_index.intra-operator | 30 |
| abstract_inverted_index.model-parallel | 10 |
| abstract_inverted_index.broadcast-based | 125 |
| abstract_inverted_index.microbenchmarks, | 134 |
| abstract_inverted_index."overlapping-friendly" | 130 |
| cited_by_percentile_year | |
| countries_distinct_count | 0 |
| institutions_distinct_count | 9 |
| sustainable_development_goals[0].id | https://metadata.un.org/sdg/9 |
| sustainable_development_goals[0].score | 0.4099999964237213 |
| sustainable_development_goals[0].display_name | Industry, innovation and infrastructure |
| citation_normalized_percentile |