{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,19]],"date-time":"2026-05-19T07:11:09Z","timestamp":1779174669577,"version":"3.51.4"},"reference-count":45,"publisher":"Association for Computing Machinery (ACM)","issue":"8","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Proc. VLDB Endow."],"published-print":{"date-parts":[[2021,4]]},"abstract":"<jats:p>\n            We consider the question: what is the abstraction that should be implemented by the computational engine of a machine learning system? Current machine learning systems typically push whole tensors through a series of compute kernels such as matrix multiplications or activation functions, where each kernel runs on an AI accelerator (ASIC) such as a GPU. This implementation abstraction provides little built-in support for ML systems to scale past a single machine, or for handling large models with matrices or tensors that do not easily fit into the RAM of an ASIC. In this paper, we present an alternative implementation abstraction called the\n            <jats:italic>tensor relational algebra<\/jats:italic>\n            (TRA). The TRA is a set-based algebra based on the relational algebra. Expressions in the TRA operate over binary tensor relations, where keys are multi-dimensional arrays and values are tensors. The TRA is easily executed with high efficiency in a parallel or distributed environment, and amenable to automatic optimization. Our empirical study shows that the optimized TRA-based back-end can significantly outperform alternatives for running ML workflows in distributed clusters.\n          <\/jats:p>","DOI":"10.14778\/3457390.3457399","type":"journal-article","created":{"date-parts":[[2021,10,21]],"date-time":"2021-10-21T22:48:38Z","timestamp":1634856518000},"page":"1338-1350","source":"Crossref","is-referenced-by-count":24,"title":["Tensor relational algebra for distributed machine learning system design"],"prefix":"10.14778","volume":"14","author":[{"given":"Binhang","family":"Yuan","sequence":"first","affiliation":[{"name":"Rice University"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Dimitrije","family":"Jankov","sequence":"additional","affiliation":[{"name":"Rice University"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jia","family":"Zou","sequence":"additional","affiliation":[{"name":"Arizona State University"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yuxin","family":"Tang","sequence":"additional","affiliation":[{"name":"Rice University"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Daniel","family":"Bourgeois","sequence":"additional","affiliation":[{"name":"Rice University"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Chris","family":"Jermaine","sequence":"additional","affiliation":[{"name":"Rice University"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"320","published-online":{"date-parts":[[2021,10,21]]},"reference":[{"key":"e_1_2_1_1_1","unstructured":"[n.d.]. Dask. https:\/\/dask.org\/.  [n.d.]. Dask. https:\/\/dask.org\/."},{"key":"e_1_2_1_2_1","unstructured":"[n.d.]. PETSc. https:\/\/www.mcs.anl.gov\/petsc\/.  [n.d.]. PETSc. https:\/\/www.mcs.anl.gov\/petsc\/."},{"key":"e_1_2_1_3_1","unstructured":"[n.d.]. PyTorch. https:\/\/pytorch.org\/.  [n.d.]. PyTorch. https:\/\/pytorch.org\/."},{"key":"e_1_2_1_4_1","unstructured":"[n.d.]. TensorFlow. https:\/\/www.tensorflow.org\/.  [n.d.]. TensorFlow. https:\/\/www.tensorflow.org\/."},{"key":"e_1_2_1_5_1","doi-asserted-by":"publisher","DOI":"10.5555\/3026877.3026899"},{"key":"e_1_2_1_6_1","doi-asserted-by":"publisher","DOI":"10.1109\/ICDE.2018.00048"},{"key":"e_1_2_1_7_1","doi-asserted-by":"publisher","DOI":"10.1145\/3196959.3196960"},{"key":"e_1_2_1_8_1","doi-asserted-by":"publisher","DOI":"10.1145\/2902251.2902280"},{"key":"e_1_2_1_9_1","doi-asserted-by":"publisher","DOI":"10.1147\/rd.395.0575"},{"key":"e_1_2_1_10_1","doi-asserted-by":"crossref","unstructured":"Satish Balay Shrirang Abhyankar Mark Adams Jed Brown Peter Brune Kris Buschelman Lisandro Dalcin Alp Dener Victor Eijkhout W Gropp etal 2019. PETSc users manual. (2019).  Satish Balay Shrirang Abhyankar Mark Adams Jed Brown Peter Brune Kris Buschelman Lisandro Dalcin Alp Dener Victor Eijkhout W Gropp et al. 2019. PETSc users manual. (2019).","DOI":"10.2172\/1614847"},{"key":"e_1_2_1_11_1","doi-asserted-by":"publisher","DOI":"10.1145\/3329486.3329498"},{"key":"e_1_2_1_12_1","volume-title":"Workshop: High Performance Computing on Stampede","volume":"262","author":"Barker Brandon","year":"2015"},{"key":"e_1_2_1_13_1","doi-asserted-by":"publisher","DOI":"10.1145\/276305.276386"},{"key":"e_1_2_1_14_1","doi-asserted-by":"publisher","DOI":"10.1145\/1807167.1807271"},{"key":"e_1_2_1_15_1","doi-asserted-by":"publisher","DOI":"10.1145\/2463676.2465283"},{"key":"e_1_2_1_16_1","doi-asserted-by":"publisher","DOI":"10.5555\/3291168.3291211"},{"key":"e_1_2_1_17_1","volume-title":"The Fourth Symposium on the Frontiers of Massively Parallel Computation. IEEE Computer Society, 120--121","author":"Choi Jaeyoung","year":"1992"},{"key":"e_1_2_1_18_1","doi-asserted-by":"publisher","DOI":"10.1145\/3357223.3362707"},{"key":"e_1_2_1_19_1","doi-asserted-by":"publisher","DOI":"10.5555\/1792734.1792766"},{"key":"e_1_2_1_20_1","doi-asserted-by":"publisher","DOI":"10.1145\/3318464.3389747"},{"key":"e_1_2_1_21_1","doi-asserted-by":"publisher","DOI":"10.1109\/ICDE.2011.5767930"},{"key":"e_1_2_1_22_1","doi-asserted-by":"publisher","DOI":"10.1109\/TPDS.2017.2686384"},{"key":"e_1_2_1_23_1","doi-asserted-by":"publisher","DOI":"10.1145\/3299869.3319865"},{"key":"e_1_2_1_24_1","volume-title":"Tensorframes on google's tensorflow and apache spark. Bay Area Spark Meetup","author":"Hunter T","year":"2016"},{"key":"e_1_2_1_25_1","doi-asserted-by":"publisher","DOI":"10.1145\/3070607.3070608"},{"key":"e_1_2_1_26_1","doi-asserted-by":"publisher","DOI":"10.14778\/3317315.3317323"},{"key":"e_1_2_1_27_1","doi-asserted-by":"publisher","DOI":"10.1145\/3318464.3380575"},{"key":"e_1_2_1_28_1","doi-asserted-by":"publisher","DOI":"10.1145\/3209889.3209896"},{"key":"e_1_2_1_29_1","doi-asserted-by":"publisher","DOI":"10.1145\/2661829.2661864"},{"key":"e_1_2_1_30_1","doi-asserted-by":"publisher","DOI":"10.1145\/2661829.2661842"},{"key":"e_1_2_1_31_1","doi-asserted-by":"publisher","DOI":"10.1145\/3133901"},{"key":"e_1_2_1_32_1","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v34i04.5881"},{"key":"e_1_2_1_33_1","doi-asserted-by":"publisher","DOI":"10.14778\/3415478.3415530"},{"key":"e_1_2_1_34_1","doi-asserted-by":"publisher","DOI":"10.14778\/3137765.3137812"},{"key":"e_1_2_1_35_1","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2018.2827988"},{"key":"e_1_2_1_36_1","doi-asserted-by":"publisher","DOI":"10.1145\/2783258.2783381"},{"key":"e_1_2_1_37_1","doi-asserted-by":"publisher","DOI":"10.1145\/2766462.2767755"},{"key":"e_1_2_1_38_1","volume-title":"Sparknet: Training deep networks in spark. arXiv preprint arXiv:1511.06051","author":"Moritz Philipp","year":"2015"},{"key":"e_1_2_1_39_1","doi-asserted-by":"publisher","DOI":"10.25080\/Majora-7b98e3ed-013"},{"key":"e_1_2_1_40_1","doi-asserted-by":"publisher","DOI":"10.5555\/2033408.2033420"},{"key":"e_1_2_1_41_1","first-page":"2","article-title":"Deeplearning4j: Open-source distributed deep learning for the jvm","volume":"2","author":"D Team","year":"2016","journal-title":"Apache Software Foundation License"},{"key":"e_1_2_1_42_1","doi-asserted-by":"publisher","DOI":"10.1002\/(SICI)1096-9128(199704)9:4<255::AID-CPE250>3.0.CO;2-2"},{"key":"e_1_2_1_43_1","volume-title":"Tensor comprehensions: Framework-agnostic high-performance machine learning abstractions. arXiv preprint arXiv:1802.04730","author":"Vasilache Nicolas","year":"2018"},{"key":"e_1_2_1_44_1","volume-title":"Speech commands: A dataset for limited-vocabulary speech recognition. arXiv preprint arXiv:1804.03209","author":"Warden Pete","year":"2018"},{"key":"e_1_2_1_45_1","doi-asserted-by":"publisher","DOI":"10.5555\/1863103.1863113"}],"container-title":["Proceedings of the VLDB Endowment"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.14778\/3457390.3457399","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,12,28]],"date-time":"2022-12-28T10:47:40Z","timestamp":1672224460000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.14778\/3457390.3457399"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021,4]]},"references-count":45,"journal-issue":{"issue":"8","published-print":{"date-parts":[[2021,4]]}},"alternative-id":["10.14778\/3457390.3457399"],"URL":"https:\/\/doi.org\/10.14778\/3457390.3457399","relation":{},"ISSN":["2150-8097"],"issn-type":[{"value":"2150-8097","type":"print"}],"subject":[],"published":{"date-parts":[[2021,4]]}}}