{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:7GICMBKPKHSEVQLGS2R3HYKZUQ","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"94561c2b1ee1a2655a3a58474127f736e5293145398cb29b20614c43e5cf1f12","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2025-10-01T15:01:29Z","title_canon_sha256":"95f98d2871d614a9b9eba44491a98ad7848df8494c7b640f9c4bf0c91f9dd44c"},"schema_version":"1.0","source":{"id":"2510.00991","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2510.00991","created_at":"2026-06-02T02:04:09Z"},{"alias_kind":"arxiv_version","alias_value":"2510.00991v2","created_at":"2026-06-02T02:04:09Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2510.00991","created_at":"2026-06-02T02:04:09Z"},{"alias_kind":"pith_short_12","alias_value":"7GICMBKPKHSE","created_at":"2026-06-02T02:04:09Z"},{"alias_kind":"pith_short_16","alias_value":"7GICMBKPKHSEVQLG","created_at":"2026-06-02T02:04:09Z"},{"alias_kind":"pith_short_8","alias_value":"7GICMBKP","created_at":"2026-06-02T02:04:09Z"}],"graph_snapshots":[{"event_id":"sha256:43191a7b3cc70b75672938a7c3db643c1773b681f6900b858a29bdc93da01172","target":"graph","created_at":"2026-06-02T02:04:09Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2510.00991/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large-scale LLM training requires collective communication libraries to exchange data among distributed GPUs. As a company dedicated to building and operating large-scale GPU training clusters, we encounter several practical limitations of NCCL in production, including 1) SM competition between computation and communication, 2) expensive restart costs under link failures, and 3) insufficient observability of transient collective communication anomalies. To address these challenges, we propose VCCL, an efficient, reliable, and observable collective communication library in large-scale GPU train","authors_text":"(2) Beihang University, 3), (3) Shanghai Innovation Institute, 4), (4) Tsinghua University, 5), (5) Zhipu AI, (6) China Unicom Research Institute, (7) Shanghai AI Power Technology Co., Aohan Zeng (4, Bohua Xu (6), Bowen Han (6), Chunming Hu (2) ((1) Infrawaves, Da Liu (1), Fangzheng Jiao (2), He Liu (1), Hongzhou Zhang (7), Jinlong Hou (3), Jun Chen (1), Ltd), Menghao Zhang (2), Mingjun Zhang (1), Mingwei Xu (4), Qing Chen (1), Ruya Gu (1), Shuaixing Duan (5), Wei Chen (1), Wenqi Xie (1), Xiaohe Hu (1, Yang Jing (1), Yanmin Jia (1), Yan Zhang (1), Yuan Cheng (3), Ziteng Chen (1","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2025-10-01T15:01:29Z","title":"An Efficient, Reliable and Observable Collective Communication Library in Large-scale GPU Training Clusters"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2510.00991","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:978724e4e3d56836511f4a44225b468802f7c2b9c83d0adef1c2f41e72e3d2da","target":"record","created_at":"2026-06-02T02:04:09Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"94561c2b1ee1a2655a3a58474127f736e5293145398cb29b20614c43e5cf1f12","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2025-10-01T15:01:29Z","title_canon_sha256":"95f98d2871d614a9b9eba44491a98ad7848df8494c7b640f9c4bf0c91f9dd44c"},"schema_version":"1.0","source":{"id":"2510.00991","kind":"arxiv","version":2}},"canonical_sha256":"f99026054f51e44ac16696a3b3e159a43a179844a7512b1b6e003f3707c47b16","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f99026054f51e44ac16696a3b3e159a43a179844a7512b1b6e003f3707c47b16","first_computed_at":"2026-06-02T02:04:09.475025Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-02T02:04:09.475025Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"JwMZGVplHppsRm3pvhCexOb951Rxne4u6R7q5XkUPtWidJpKbNwZlHy9hRME/n59doNQfLYuViD1UmRT+PJTBg==","signature_status":"signed_v1","signed_at":"2026-06-02T02:04:09.475622Z","signed_message":"canonical_sha256_bytes"},"source_id":"2510.00991","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:978724e4e3d56836511f4a44225b468802f7c2b9c83d0adef1c2f41e72e3d2da","sha256:43191a7b3cc70b75672938a7c3db643c1773b681f6900b858a29bdc93da01172"],"state_sha256":"5ae44585ade29c7b4a69f9811a64f3dfe5af7c2eb7c2bb61d3aa09d47faef4a2"}