{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:MN7JPC5HVYTWLHJV72SQHGYWIB","short_pith_number":"pith:MN7JPC5H","schema_version":"1.0","canonical_sha256":"637e978ba7ae27659d35fea5039b16405accf565dcc7792b2c6d01ac3ab9d916","source":{"kind":"arxiv","id":"2606.18741","version":1},"attestation_state":"computed","paper":{"title":"ReMP: Low-Downtime Runtime Model-Parallelism Reconfiguration for LLM Serving","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.DC","authors_text":"Baodong Wu, Daning Cheng, Haipeng Yuan, Kaining Zheng, Xiang Gao, Yongshu Bai, Yuchen Zhang, Yunquan Zhang","submitted_at":"2026-06-17T06:36:40Z","abstract_excerpt":"Current large language model (LLM) inference systems universally deploy ultra-large-scale models using a combination of Tensor Parallelism (TP) and Pipeline Parallelism (PP). However, existing systems treat the model parallelism topology as a static configuration that cannot be flexibly adjusted at runtime. This rigid design creates a fundamental contradiction with the dynamically changing inference workloads in real-world scenarios. State-of-the-art systems lack online reconfiguration capabilities and can only switch configurations by restarting the service, resulting in several minutes of se"},"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":"2606.18741","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.DC","submitted_at":"2026-06-17T06:36:40Z","cross_cats_sorted":[],"title_canon_sha256":"599d5a1bbb2088896cf765c0f5a555e490dab5662eaebf2ebd9dad693d5b7fc8","abstract_canon_sha256":"69c94ed17da1abc61bd1c111c860498cc6915c39908ad7cb63d7447c46023e10"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-19T16:11:46.173164Z","signature_b64":"rQNf8GnPuIqetCAD/e8NXipNb07NOB7JDnczFjA3LfZbdPYFamljcK4p+YnkK172MkpFeFDBRKRfEm4E+tXoAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"637e978ba7ae27659d35fea5039b16405accf565dcc7792b2c6d01ac3ab9d916","last_reissued_at":"2026-06-19T16:11:46.172824Z","signature_status":"signed_v1","first_computed_at":"2026-06-19T16:11:46.172824Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ReMP: Low-Downtime Runtime Model-Parallelism Reconfiguration for LLM Serving","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.DC","authors_text":"Baodong Wu, Daning Cheng, Haipeng Yuan, Kaining Zheng, Xiang Gao, Yongshu Bai, Yuchen Zhang, Yunquan Zhang","submitted_at":"2026-06-17T06:36:40Z","abstract_excerpt":"Current large language model (LLM) inference systems universally deploy ultra-large-scale models using a combination of Tensor Parallelism (TP) and Pipeline Parallelism (PP). However, existing systems treat the model parallelism topology as a static configuration that cannot be flexibly adjusted at runtime. This rigid design creates a fundamental contradiction with the dynamically changing inference workloads in real-world scenarios. State-of-the-art systems lack online reconfiguration capabilities and can only switch configurations by restarting the service, resulting in several minutes of se"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.18741","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/2606.18741/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":"2606.18741","created_at":"2026-06-19T16:11:46.172885+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.18741v1","created_at":"2026-06-19T16:11:46.172885+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.18741","created_at":"2026-06-19T16:11:46.172885+00:00"},{"alias_kind":"pith_short_12","alias_value":"MN7JPC5HVYTW","created_at":"2026-06-19T16:11:46.172885+00:00"},{"alias_kind":"pith_short_16","alias_value":"MN7JPC5HVYTWLHJV","created_at":"2026-06-19T16:11:46.172885+00:00"},{"alias_kind":"pith_short_8","alias_value":"MN7JPC5H","created_at":"2026-06-19T16:11:46.172885+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/MN7JPC5HVYTWLHJV72SQHGYWIB","json":"https://pith.science/pith/MN7JPC5HVYTWLHJV72SQHGYWIB.json","graph_json":"https://pith.science/api/pith-number/MN7JPC5HVYTWLHJV72SQHGYWIB/graph.json","events_json":"https://pith.science/api/pith-number/MN7JPC5HVYTWLHJV72SQHGYWIB/events.json","paper":"https://pith.science/paper/MN7JPC5H"},"agent_actions":{"view_html":"https://pith.science/pith/MN7JPC5HVYTWLHJV72SQHGYWIB","download_json":"https://pith.science/pith/MN7JPC5HVYTWLHJV72SQHGYWIB.json","view_paper":"https://pith.science/paper/MN7JPC5H","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.18741&json=true","fetch_graph":"https://pith.science/api/pith-number/MN7JPC5HVYTWLHJV72SQHGYWIB/graph.json","fetch_events":"https://pith.science/api/pith-number/MN7JPC5HVYTWLHJV72SQHGYWIB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MN7JPC5HVYTWLHJV72SQHGYWIB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MN7JPC5HVYTWLHJV72SQHGYWIB/action/storage_attestation","attest_author":"https://pith.science/pith/MN7JPC5HVYTWLHJV72SQHGYWIB/action/author_attestation","sign_citation":"https://pith.science/pith/MN7JPC5HVYTWLHJV72SQHGYWIB/action/citation_signature","submit_replication":"https://pith.science/pith/MN7JPC5HVYTWLHJV72SQHGYWIB/action/replication_record"}},"created_at":"2026-06-19T16:11:46.172885+00:00","updated_at":"2026-06-19T16:11:46.172885+00:00"}