{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:ODQ6JKCNE32IJKBW4SPKZJVOL7","short_pith_number":"pith:ODQ6JKCN","schema_version":"1.0","canonical_sha256":"70e1e4a84d26f484a836e49eaca6ae5fd042c0bc1714250b5cb28aed67e9d070","source":{"kind":"arxiv","id":"2607.12121","version":1},"attestation_state":"computed","paper":{"title":"FlashDiff: Efficient Regional Execution and Scheduling for Diffusion Model Serving","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.DC","authors_text":"Ayush Barik, Chensong Zhang, Fan Lai, Ping He, Songrun Xie, Yaqi Qiao, Zhengzhong Tu","submitted_at":"2026-07-13T20:00:39Z","abstract_excerpt":"Diffusion models have become the central backbone for modern image, video, and audio generation, but their efficient service remains a challenge. Unlike autoregressive decoding, diffusion inference repeatedly updates high-dimensional spatial or temporal latents over many denoising steps. This all-region execution pattern makes generation latency high and limits serving throughput. Existing multi-GPU parallelization methods can reduce per-step computation, but often introduce substantial activation exchange overhead, causing communication to offset or even outweigh the benefits of parallel exec"},"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":"2607.12121","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2026-07-13T20:00:39Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"3a47cdc15f4ca5d4463f6a36349a22b057c78ba2673ef0916558d21b27e8ef59","abstract_canon_sha256":"fb139bd85d808dc0c09756e3eaf5db98e7e6e39912a86a66c12b11a07d943c58"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-15T00:21:34.449599Z","signature_b64":"6936Sgr/O86FMYuC8h7yDnjQKjXh85Mz9b3fw5HbCVksnumtLgMgjsKWrZrrPBJRl9EfnNGrL57FBDoqKkcFBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"70e1e4a84d26f484a836e49eaca6ae5fd042c0bc1714250b5cb28aed67e9d070","last_reissued_at":"2026-07-15T00:21:34.448764Z","signature_status":"signed_v1","first_computed_at":"2026-07-15T00:21:34.448764Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"FlashDiff: Efficient Regional Execution and Scheduling for Diffusion Model Serving","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.DC","authors_text":"Ayush Barik, Chensong Zhang, Fan Lai, Ping He, Songrun Xie, Yaqi Qiao, Zhengzhong Tu","submitted_at":"2026-07-13T20:00:39Z","abstract_excerpt":"Diffusion models have become the central backbone for modern image, video, and audio generation, but their efficient service remains a challenge. Unlike autoregressive decoding, diffusion inference repeatedly updates high-dimensional spatial or temporal latents over many denoising steps. This all-region execution pattern makes generation latency high and limits serving throughput. Existing multi-GPU parallelization methods can reduce per-step computation, but often introduce substantial activation exchange overhead, causing communication to offset or even outweigh the benefits of parallel exec"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.12121","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/2607.12121/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":"2607.12121","created_at":"2026-07-15T00:21:34.449212+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.12121v1","created_at":"2026-07-15T00:21:34.449212+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.12121","created_at":"2026-07-15T00:21:34.449212+00:00"},{"alias_kind":"pith_short_12","alias_value":"ODQ6JKCNE32I","created_at":"2026-07-15T00:21:34.449212+00:00"},{"alias_kind":"pith_short_16","alias_value":"ODQ6JKCNE32IJKBW","created_at":"2026-07-15T00:21:34.449212+00:00"},{"alias_kind":"pith_short_8","alias_value":"ODQ6JKCN","created_at":"2026-07-15T00:21:34.449212+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/ODQ6JKCNE32IJKBW4SPKZJVOL7","json":"https://pith.science/pith/ODQ6JKCNE32IJKBW4SPKZJVOL7.json","graph_json":"https://pith.science/api/pith-number/ODQ6JKCNE32IJKBW4SPKZJVOL7/graph.json","events_json":"https://pith.science/api/pith-number/ODQ6JKCNE32IJKBW4SPKZJVOL7/events.json","paper":"https://pith.science/paper/ODQ6JKCN"},"agent_actions":{"view_html":"https://pith.science/pith/ODQ6JKCNE32IJKBW4SPKZJVOL7","download_json":"https://pith.science/pith/ODQ6JKCNE32IJKBW4SPKZJVOL7.json","view_paper":"https://pith.science/paper/ODQ6JKCN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.12121&json=true","fetch_graph":"https://pith.science/api/pith-number/ODQ6JKCNE32IJKBW4SPKZJVOL7/graph.json","fetch_events":"https://pith.science/api/pith-number/ODQ6JKCNE32IJKBW4SPKZJVOL7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ODQ6JKCNE32IJKBW4SPKZJVOL7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ODQ6JKCNE32IJKBW4SPKZJVOL7/action/storage_attestation","attest_author":"https://pith.science/pith/ODQ6JKCNE32IJKBW4SPKZJVOL7/action/author_attestation","sign_citation":"https://pith.science/pith/ODQ6JKCNE32IJKBW4SPKZJVOL7/action/citation_signature","submit_replication":"https://pith.science/pith/ODQ6JKCNE32IJKBW4SPKZJVOL7/action/replication_record"}},"created_at":"2026-07-15T00:21:34.449212+00:00","updated_at":"2026-07-15T00:21:34.449212+00:00"}