{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:Z5C2UCO7VZKO4BBTFM7G6KCDIP","short_pith_number":"pith:Z5C2UCO7","schema_version":"1.0","canonical_sha256":"cf45aa09dfae54ee04332b3e6f284343eba24067b842bdae4a7b6834dd8bf3c2","source":{"kind":"arxiv","id":"2507.11830","version":1},"attestation_state":"computed","paper":{"title":"Arctic Inference with Shift Parallelism: Fast and Efficient Open Source Inference System for Enterprise AI","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.DC","authors_text":"Aurick Qiao, Jeff Rasley, Juncheng Yang, Mert Hidayetoglu, Michael Wyatt, Samyam Rajbhandari, Ye Wang, Yuxiong He","submitted_at":"2025-07-16T01:32:31Z","abstract_excerpt":"Inference is now the dominant AI workload, yet existing systems force trade-offs between latency, throughput, and cost. Arctic Inference, an open-source vLLM plugin from Snowflake AI Research, introduces Shift Parallelism, a dynamic parallelism strategy that adapts to real-world traffic while integrating speculative decoding, SwiftKV compute reduction, and optimized embedding inference. It achieves up to 3.4 times faster request completion, 1.75 times faster generation, and 1.6M tokens/sec per GPU for embeddings, outperforming both latency- and throughput-optimized deployments. Already powerin"},"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":"2507.11830","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2025-07-16T01:32:31Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"9684e498fb62c3c491ebcde34b48a502cd66d7d56f36abcecb9dd861cde5c3c9","abstract_canon_sha256":"32baeeb6922a57bae1d4ce0d48508045ab9661ccfb8922e57961db1e7061ccff"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:37:51.658657Z","signature_b64":"LZLcjKPxlRvgrpLjeiXrFre+WnWwHTuQZkkEBWG8wQHNIItlbuUz2gios6XXUUQAnyS67TFsapAc1dfYckrkBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cf45aa09dfae54ee04332b3e6f284343eba24067b842bdae4a7b6834dd8bf3c2","last_reissued_at":"2026-07-05T11:37:51.658165Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:37:51.658165Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Arctic Inference with Shift Parallelism: Fast and Efficient Open Source Inference System for Enterprise AI","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.DC","authors_text":"Aurick Qiao, Jeff Rasley, Juncheng Yang, Mert Hidayetoglu, Michael Wyatt, Samyam Rajbhandari, Ye Wang, Yuxiong He","submitted_at":"2025-07-16T01:32:31Z","abstract_excerpt":"Inference is now the dominant AI workload, yet existing systems force trade-offs between latency, throughput, and cost. Arctic Inference, an open-source vLLM plugin from Snowflake AI Research, introduces Shift Parallelism, a dynamic parallelism strategy that adapts to real-world traffic while integrating speculative decoding, SwiftKV compute reduction, and optimized embedding inference. It achieves up to 3.4 times faster request completion, 1.75 times faster generation, and 1.6M tokens/sec per GPU for embeddings, outperforming both latency- and throughput-optimized deployments. Already powerin"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.11830","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/2507.11830/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":"2507.11830","created_at":"2026-07-05T11:37:51.658230+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.11830v1","created_at":"2026-07-05T11:37:51.658230+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.11830","created_at":"2026-07-05T11:37:51.658230+00:00"},{"alias_kind":"pith_short_12","alias_value":"Z5C2UCO7VZKO","created_at":"2026-07-05T11:37:51.658230+00:00"},{"alias_kind":"pith_short_16","alias_value":"Z5C2UCO7VZKO4BBT","created_at":"2026-07-05T11:37:51.658230+00:00"},{"alias_kind":"pith_short_8","alias_value":"Z5C2UCO7","created_at":"2026-07-05T11:37:51.658230+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/Z5C2UCO7VZKO4BBTFM7G6KCDIP","json":"https://pith.science/pith/Z5C2UCO7VZKO4BBTFM7G6KCDIP.json","graph_json":"https://pith.science/api/pith-number/Z5C2UCO7VZKO4BBTFM7G6KCDIP/graph.json","events_json":"https://pith.science/api/pith-number/Z5C2UCO7VZKO4BBTFM7G6KCDIP/events.json","paper":"https://pith.science/paper/Z5C2UCO7"},"agent_actions":{"view_html":"https://pith.science/pith/Z5C2UCO7VZKO4BBTFM7G6KCDIP","download_json":"https://pith.science/pith/Z5C2UCO7VZKO4BBTFM7G6KCDIP.json","view_paper":"https://pith.science/paper/Z5C2UCO7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.11830&json=true","fetch_graph":"https://pith.science/api/pith-number/Z5C2UCO7VZKO4BBTFM7G6KCDIP/graph.json","fetch_events":"https://pith.science/api/pith-number/Z5C2UCO7VZKO4BBTFM7G6KCDIP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/Z5C2UCO7VZKO4BBTFM7G6KCDIP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/Z5C2UCO7VZKO4BBTFM7G6KCDIP/action/storage_attestation","attest_author":"https://pith.science/pith/Z5C2UCO7VZKO4BBTFM7G6KCDIP/action/author_attestation","sign_citation":"https://pith.science/pith/Z5C2UCO7VZKO4BBTFM7G6KCDIP/action/citation_signature","submit_replication":"https://pith.science/pith/Z5C2UCO7VZKO4BBTFM7G6KCDIP/action/replication_record"}},"created_at":"2026-07-05T11:37:51.658230+00:00","updated_at":"2026-07-05T11:37:51.658230+00:00"}