{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:KAU2XWEATOF7FOMSACCLMKPFQW","short_pith_number":"pith:KAU2XWEA","schema_version":"1.0","canonical_sha256":"5029abd8809b8bf2b9920084b629e585bdc14d617393fbf0809e796c601e80a5","source":{"kind":"arxiv","id":"2411.01738","version":1},"attestation_state":"computed","paper":{"title":"xDiT: an Inference Engine for Diffusion Transformers (DiTs) with Massive Parallelism","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.DC","authors_text":"Aoyu Li, Jiannan Wang, Jiarui Fang, Jinzhe Pan, Xibo Sun","submitted_at":"2024-11-04T01:40:38Z","abstract_excerpt":"Diffusion models are pivotal for generating high-quality images and videos. Inspired by the success of OpenAI's Sora, the backbone of diffusion models is evolving from U-Net to Transformer, known as Diffusion Transformers (DiTs). However, generating high-quality content necessitates longer sequence lengths, exponentially increasing the computation required for the attention mechanism, and escalating DiTs inference latency. Parallel inference is essential for real-time DiTs deployments, but relying on a single parallel method is impractical due to poor scalability at large scales. This paper in"},"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":"2411.01738","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2024-11-04T01:40:38Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"92edd2f35cc467049fdbb6ceb46cca36ed4760b858672b1166e40bfe73671ad6","abstract_canon_sha256":"b6924bc544d567c0abdf06bc33c6f812cbc5c962dc4b8a6ee33bc0c457016761"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:30:35.585834Z","signature_b64":"TkSL2xUZrS+f0WI5RR7TlcCKzC0F/xESbQSHyWFMm+ifJnMQgIuSk1xbbVsLOApIU9i1Uu3CR810paWWqRiVBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5029abd8809b8bf2b9920084b629e585bdc14d617393fbf0809e796c601e80a5","last_reissued_at":"2026-07-05T09:30:35.585440Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:30:35.585440Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"xDiT: an Inference Engine for Diffusion Transformers (DiTs) with Massive Parallelism","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.DC","authors_text":"Aoyu Li, Jiannan Wang, Jiarui Fang, Jinzhe Pan, Xibo Sun","submitted_at":"2024-11-04T01:40:38Z","abstract_excerpt":"Diffusion models are pivotal for generating high-quality images and videos. Inspired by the success of OpenAI's Sora, the backbone of diffusion models is evolving from U-Net to Transformer, known as Diffusion Transformers (DiTs). However, generating high-quality content necessitates longer sequence lengths, exponentially increasing the computation required for the attention mechanism, and escalating DiTs inference latency. Parallel inference is essential for real-time DiTs deployments, but relying on a single parallel method is impractical due to poor scalability at large scales. This paper in"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.01738","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/2411.01738/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":"2411.01738","created_at":"2026-07-05T09:30:35.585497+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.01738v1","created_at":"2026-07-05T09:30:35.585497+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.01738","created_at":"2026-07-05T09:30:35.585497+00:00"},{"alias_kind":"pith_short_12","alias_value":"KAU2XWEATOF7","created_at":"2026-07-05T09:30:35.585497+00:00"},{"alias_kind":"pith_short_16","alias_value":"KAU2XWEATOF7FOMS","created_at":"2026-07-05T09:30:35.585497+00:00"},{"alias_kind":"pith_short_8","alias_value":"KAU2XWEA","created_at":"2026-07-05T09:30:35.585497+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":16,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.07534","citing_title":"Infinite Worlds with Versatile Interactions","ref_index":13,"is_internal_anchor":true},{"citing_arxiv_id":"2606.19271","citing_title":"TurboServe: Serving Streaming Video Generation Efficiently and Economically","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2606.13501","citing_title":"GF-DiT: Scheduling Parallelism for Diffusion Transformer Serving","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2606.17566","citing_title":"AoiZora: Topology-Aware Auto-Parallel Optimization for Inference of Diffusion Transformers","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2606.13501","citing_title":"GF-DiT: Scheduling Parallelism for Diffusion Transformer Serving","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2606.12688","citing_title":"M*: A Modular, Extensible, Serving System for Multimodal Models","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2511.20714","citing_title":"Inferix: A Block-Diffusion based Next-Generation Inference Engine for World Simulation","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2512.14098","citing_title":"Cornfigurator: Automated Planning for Any-to-Any Multimodal Model Serving","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2603.12118","citing_title":"Cornserve: A Distributed Serving System for Any-to-Any Multimodal Models","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2603.28489","citing_title":"Video Generation Models as World Models: Efficient Paradigms, Architectures and Algorithms","ref_index":142,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11335","citing_title":"ChunkFlow: Communication-Aware Chunked Prefetching for Layerwise Offloading in Distributed Diffusion Transformer Inference","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24447","citing_title":"Characterizing Vision-Language-Action Models across XPUs: Constraints and Acceleration for On-Robot Deployment","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2604.08123","citing_title":"LegoDiffusion: Micro-Serving Text-to-Image Diffusion Workflows","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04335","citing_title":"GENSERVE: Efficient Co-Serving of Heterogeneous Diffusion Model Workloads","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14561","citing_title":"CoCoDiff: Optimizing Collective Communications for Distributed Diffusion Transformer Inference Under Ulysses Sequence Parallelism","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15911","citing_title":"Efficient Video Diffusion Models: Advancements and Challenges","ref_index":262,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KAU2XWEATOF7FOMSACCLMKPFQW","json":"https://pith.science/pith/KAU2XWEATOF7FOMSACCLMKPFQW.json","graph_json":"https://pith.science/api/pith-number/KAU2XWEATOF7FOMSACCLMKPFQW/graph.json","events_json":"https://pith.science/api/pith-number/KAU2XWEATOF7FOMSACCLMKPFQW/events.json","paper":"https://pith.science/paper/KAU2XWEA"},"agent_actions":{"view_html":"https://pith.science/pith/KAU2XWEATOF7FOMSACCLMKPFQW","download_json":"https://pith.science/pith/KAU2XWEATOF7FOMSACCLMKPFQW.json","view_paper":"https://pith.science/paper/KAU2XWEA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.01738&json=true","fetch_graph":"https://pith.science/api/pith-number/KAU2XWEATOF7FOMSACCLMKPFQW/graph.json","fetch_events":"https://pith.science/api/pith-number/KAU2XWEATOF7FOMSACCLMKPFQW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KAU2XWEATOF7FOMSACCLMKPFQW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KAU2XWEATOF7FOMSACCLMKPFQW/action/storage_attestation","attest_author":"https://pith.science/pith/KAU2XWEATOF7FOMSACCLMKPFQW/action/author_attestation","sign_citation":"https://pith.science/pith/KAU2XWEATOF7FOMSACCLMKPFQW/action/citation_signature","submit_replication":"https://pith.science/pith/KAU2XWEATOF7FOMSACCLMKPFQW/action/replication_record"}},"created_at":"2026-07-05T09:30:35.585497+00:00","updated_at":"2026-07-05T09:30:35.585497+00:00"}