{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:3PFMK7CCAK5E377H5T7LCGQXOT","short_pith_number":"pith:3PFMK7CC","schema_version":"1.0","canonical_sha256":"dbcac57c4202ba4dffe7ecfeb11a1774c2a318e7f56f752adc3a038ac9719ae3","source":{"kind":"arxiv","id":"2606.01680","version":1},"attestation_state":"computed","paper":{"title":"Don't Let a Few Network Failures Slow the Entire AllReduce","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","cs.NI"],"primary_cat":"cs.DC","authors_text":"Jiedong Jiang, Nengneng Yu, Peiqing Chen, Sixian Xiong, Wei Wang, Yuefeng Wang, Zaoxing Liu","submitted_at":"2026-06-01T04:40:41Z","abstract_excerpt":"Network failures are among the most frequent hardware faults in large-scale GPU clusters and a leading cause of training-job interruptions. Modern collective communication libraries such as NCCL mitigate network failures by rerouting traffic through surviving NICs on the same server, trading reduced inter-node bandwidth for uninterrupted training. However, the degraded server remains on the critical path of the standard ring algorithm, slowing the entire collective. We present the first information-theoretic lower bound on AllReduce completion time under asymmetric network bandwidth and show t"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2606.01680","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2026-06-01T04:40:41Z","cross_cats_sorted":["cs.LG","cs.NI"],"title_canon_sha256":"acca9c2a4c9d8d1bc54fa537fef94b5ad89df7618a84a001d3d148a4bca43a63","abstract_canon_sha256":"7c430029d15ecfb4fd449e8bfe0da239f8b792af5255a2d8cca84b7006e529e2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T02:04:39.963366Z","signature_b64":"6S8mH+YFqK1d1ozh/YFhf3SpPm1U3PD7V2o+yvUIUdpodmyMV96R+2f0/L3KpJkgYfrdxHgJblU2/hCjLL73BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dbcac57c4202ba4dffe7ecfeb11a1774c2a318e7f56f752adc3a038ac9719ae3","last_reissued_at":"2026-06-02T02:04:39.963028Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T02:04:39.963028Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Don't Let a Few Network Failures Slow the Entire AllReduce","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","cs.NI"],"primary_cat":"cs.DC","authors_text":"Jiedong Jiang, Nengneng Yu, Peiqing Chen, Sixian Xiong, Wei Wang, Yuefeng Wang, Zaoxing Liu","submitted_at":"2026-06-01T04:40:41Z","abstract_excerpt":"Network failures are among the most frequent hardware faults in large-scale GPU clusters and a leading cause of training-job interruptions. Modern collective communication libraries such as NCCL mitigate network failures by rerouting traffic through surviving NICs on the same server, trading reduced inter-node bandwidth for uninterrupted training. However, the degraded server remains on the critical path of the standard ring algorithm, slowing the entire collective. We present the first information-theoretic lower bound on AllReduce completion time under asymmetric network bandwidth and show t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.01680","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2606.01680/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2606.01680","created_at":"2026-06-02T02:04:39.963077+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.01680v1","created_at":"2026-06-02T02:04:39.963077+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.01680","created_at":"2026-06-02T02:04:39.963077+00:00"},{"alias_kind":"pith_short_12","alias_value":"3PFMK7CCAK5E","created_at":"2026-06-02T02:04:39.963077+00:00"},{"alias_kind":"pith_short_16","alias_value":"3PFMK7CCAK5E377H","created_at":"2026-06-02T02:04:39.963077+00:00"},{"alias_kind":"pith_short_8","alias_value":"3PFMK7CC","created_at":"2026-06-02T02:04:39.963077+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3PFMK7CCAK5E377H5T7LCGQXOT","json":"https://pith.science/pith/3PFMK7CCAK5E377H5T7LCGQXOT.json","graph_json":"https://pith.science/api/pith-number/3PFMK7CCAK5E377H5T7LCGQXOT/graph.json","events_json":"https://pith.science/api/pith-number/3PFMK7CCAK5E377H5T7LCGQXOT/events.json","paper":"https://pith.science/paper/3PFMK7CC"},"agent_actions":{"view_html":"https://pith.science/pith/3PFMK7CCAK5E377H5T7LCGQXOT","download_json":"https://pith.science/pith/3PFMK7CCAK5E377H5T7LCGQXOT.json","view_paper":"https://pith.science/paper/3PFMK7CC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.01680&json=true","fetch_graph":"https://pith.science/api/pith-number/3PFMK7CCAK5E377H5T7LCGQXOT/graph.json","fetch_events":"https://pith.science/api/pith-number/3PFMK7CCAK5E377H5T7LCGQXOT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3PFMK7CCAK5E377H5T7LCGQXOT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3PFMK7CCAK5E377H5T7LCGQXOT/action/storage_attestation","attest_author":"https://pith.science/pith/3PFMK7CCAK5E377H5T7LCGQXOT/action/author_attestation","sign_citation":"https://pith.science/pith/3PFMK7CCAK5E377H5T7LCGQXOT/action/citation_signature","submit_replication":"https://pith.science/pith/3PFMK7CCAK5E377H5T7LCGQXOT/action/replication_record"}},"created_at":"2026-06-02T02:04:39.963077+00:00","updated_at":"2026-06-02T02:04:39.963077+00:00"}