{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,23]],"date-time":"2026-04-23T08:00:15Z","timestamp":1776931215021,"version":"3.51.2"},"publisher-location":"New York, NY, USA","reference-count":14,"publisher":"ACM","license":[{"start":{"date-parts":[[2026,11,15]],"date-time":"2026-11-15T00:00:00Z","timestamp":1794700800000},"content-version":"vor","delay-in-days":365,"URL":"http:\/\/www.acm.org\/publications\/policies\/copyright_policy#Background"}],"funder":[{"name":"Department of Energy","award":["DE-AC05-00OR22725"],"award-info":[{"award-number":["DE-AC05-00OR22725"]}]}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2025,11,16]]},"DOI":"10.1145\/3731599.3767471","type":"proceedings-article","created":{"date-parts":[[2025,11,7]],"date-time":"2025-11-07T16:13:44Z","timestamp":1762532024000},"page":"997-1002","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["From Exploration to Explanation: ML-Driven Causal Discovery for Datacenter Reliability at Scale"],"prefix":"10.1145","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-5752-5778","authenticated-orcid":false,"given":"Pavana","family":"Prakash","sequence":"first","affiliation":[{"name":"Hewlett Packard Labs, Milpitas, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0008-5652-4408","authenticated-orcid":false,"given":"Rolando P.","family":"Hong Enriquez","sequence":"additional","affiliation":[{"name":"Hewlett Packard Labs, Oxford, United Kingdom"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6963-9337","authenticated-orcid":false,"given":"Sergey","family":"Serebryakov","sequence":"additional","affiliation":[{"name":"Hewlett Packard Labs, Milpitas, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0005-8435-6446","authenticated-orcid":false,"given":"David","family":"Grant","sequence":"additional","affiliation":[{"name":"Oak Ridge National Laboratory (ORNL), Oak Ridge, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3639-3956","authenticated-orcid":false,"given":"Wesley","family":"Brewer","sequence":"additional","affiliation":[{"name":"Oak Ridge National Laboratory (ORNL), Oak Ridge, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9830-8588","authenticated-orcid":false,"given":"Dejan","family":"Milojicic","sequence":"additional","affiliation":[{"name":"Hewlett Packard Labs, Milpitas, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"320","published-online":{"date-parts":[[2025,11,15]]},"reference":[{"key":"e_1_3_3_1_2_2","doi-asserted-by":"publisher","DOI":"10.1109\/SC41406.2024.00029"},{"key":"e_1_3_3_1_3_2","unstructured":"Brightlio. 2024. Technology Illuminated: Data Center Statistics in 2024. https:\/\/brightlio.com\/data-center-stats\/"},{"key":"e_1_3_3_1_4_2","doi-asserted-by":"crossref","unstructured":"Clive\u00a0WJ Granger. 1969. Investigating causal relations by econometric models and cross-spectral methods. Econometrica 37 3 (1969) 424\u2013438.","DOI":"10.2307\/1912791"},{"key":"e_1_3_3_1_5_2","volume-title":"The Thirteenth International Conference on Learning Representations","author":"Han Xiao","year":"2025","unstructured":"Xiao Han, Saima Absar, Lu Zhang, and Shuhan Yuan. 2025. Root Cause Analysis of Anomalies in Multivariate Time Series through Granger Causal Discovery. In The Thirteenth International Conference on Learning Representations."},{"key":"e_1_3_3_1_6_2","doi-asserted-by":"crossref","unstructured":"Martin Molan Andrea Borghesi Daniele Cesarini Luca Benini and Andrea Bartolini. 2023. RUAD: Unsupervised anomaly detection in HPC systems. Future Generation Computer Systems 141 (2023) 542\u2013554.","DOI":"10.1016\/j.future.2022.12.001"},{"key":"e_1_3_3_1_7_2","volume-title":"2025 State of Reliability: Overview","year":"2025","unstructured":"NERC. 2025. 2025 State of Reliability: Overview. Technical Report. North American Electric Reliability Corporation (NERC), Atlanta, GA. https:\/\/www.nerc.com\/pa\/RAPA\/PA\/Performance%20Analysis%20DL\/NERC_SOR_2025_Overview.pdf"},{"key":"e_1_3_3_1_8_2","doi-asserted-by":"publisher","DOI":"10.1109\/ISEMV63338.2024.00024"},{"key":"e_1_3_3_1_9_2","doi-asserted-by":"crossref","unstructured":"Jakob Runge Peer Nowack Marlene Kretschmer Seth Flaxman and Dino Sejdinovic. 2019. Detecting causal associations in large nonlinear time series datasets. Science Advances 5 11 (2019) eaau4996.","DOI":"10.1126\/sciadv.aau4996"},{"key":"e_1_3_3_1_10_2","doi-asserted-by":"publisher","DOI":"10.1145\/3458817.3476188"},{"key":"e_1_3_3_1_11_2","doi-asserted-by":"publisher","DOI":"10.1109\/SCW63240.2024.00120"},{"key":"e_1_3_3_1_12_2","doi-asserted-by":"publisher","DOI":"10.1109\/HPEC43674.2020.9286239"},{"key":"e_1_3_3_1_13_2","volume-title":"Announcing supercomputer summit","author":"Wells Jack","year":"2016","unstructured":"Jack Wells, Buddy Bland, Jeff Nichols, Jim Hack, Fernanda Foertter, Gaute Hagen, Thomas Maier, Moetasim Ashfaq, Bronson Messer, and Suzanne Parete-Koon. 2016. Announcing supercomputer summit. Technical Report. Oak Ridge National Lab.(ORNL), Oak Ridge, TN (United States)."},{"key":"e_1_3_3_1_14_2","unstructured":"Wenzhuo Yang Kun Zhang and Steven\u00a0CH Hoi. 2022. A causal approach to detecting multivariate time-series anomalies and root causes. arXiv preprint arXiv:https:\/\/arXiv.org\/abs\/2206.15033 (2022)."},{"key":"e_1_3_3_1_15_2","unstructured":"Yokogawa Electric Corporation. 2025. SMARTDAC+ Data Acquisition System. https:\/\/www.yokogawa.com\/solutions\/products-and-services\/data-acquisition\/smartdac-plus-series\/"}],"event":{"name":"SC Workshops '25: Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis","location":"St Louis MO USA","acronym":"SC Workshops '25","sponsor":["SIGHPC ACM Special Interest Group on High Performance Computing, Special Interest Group on High Performance Computing"]},"container-title":["Proceedings of the SC '25 Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis"],"original-title":[],"link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3731599.3767471","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3731599.3767471","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,9]],"date-time":"2026-01-09T19:27:29Z","timestamp":1767986849000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3731599.3767471"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,11,15]]},"references-count":14,"alternative-id":["10.1145\/3731599.3767471","10.1145\/3731599"],"URL":"https:\/\/doi.org\/10.1145\/3731599.3767471","relation":{},"subject":[],"published":{"date-parts":[[2025,11,15]]},"assertion":[{"value":"2025-11-15","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}