{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:QD3PX54SC5BQH26N6YSZY66SH4","short_pith_number":"pith:QD3PX54S","schema_version":"1.0","canonical_sha256":"80f6fbf792174303ebcdf6259c7bd23f3a38329316a109192018d9211675f3e9","source":{"kind":"arxiv","id":"2608.03741","version":1},"attestation_state":"computed","paper":{"title":"When Does Disaggregation Pay? Simulating Prefill--Decode--Attention--FFN Specialization for Agentic LLM Inference","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.DC","authors_text":"Aaron Zhao, Can Xiao, George A. Constantinides, Haoran Wu, Jiayi Nie, Przemyslaw Forys, Rika Antonova, Robert Mullins, Timothy Jones, Tony Liu, Wayne Luk","submitted_at":"2026-08-04T14:35:25Z","abstract_excerpt":"Agentic inference now dominates the LLM inference landscape, requiring LLMs to actively engage in multi-turn interactions with tool-calling capabilities. This introduces a more complex workload for the underlying inference system: serving stages such as prefill and decode exhibit substantially different behaviors and demand distinct compute and memory-bandwidth capabilities. As a result, a single homogeneous GPU system now struggles to support agentic inference, motivating an industry shift toward heterogeneous systems with disaggregated serving capabilities, such as the emerging Vera-Rubin pl"},"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":"2608.03741","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2026-08-04T14:35:25Z","cross_cats_sorted":[],"title_canon_sha256":"2796d257d3baabe8195870f64401c82d2521fc0e4ff2204636496d155b4d04b5","abstract_canon_sha256":"2fcdd3668149c8b316d6ff9d8c3fb4018a9b43efb6b8cfca61d20f055c502b83"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-05T01:37:00.152775Z","signature_b64":"MteBa8MEWjy5zszfBSI76mHT3myLmA0BkkhpNLWKGGKzB9Cmos5nfmf0yt7A6KyA5OGUHlOC23BOMOft+mqxDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"80f6fbf792174303ebcdf6259c7bd23f3a38329316a109192018d9211675f3e9","last_reissued_at":"2026-08-05T01:37:00.151238Z","signature_status":"signed_v1","first_computed_at":"2026-08-05T01:37:00.151238Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"When Does Disaggregation Pay? Simulating Prefill--Decode--Attention--FFN Specialization for Agentic LLM Inference","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.DC","authors_text":"Aaron Zhao, Can Xiao, George A. Constantinides, Haoran Wu, Jiayi Nie, Przemyslaw Forys, Rika Antonova, Robert Mullins, Timothy Jones, Tony Liu, Wayne Luk","submitted_at":"2026-08-04T14:35:25Z","abstract_excerpt":"Agentic inference now dominates the LLM inference landscape, requiring LLMs to actively engage in multi-turn interactions with tool-calling capabilities. This introduces a more complex workload for the underlying inference system: serving stages such as prefill and decode exhibit substantially different behaviors and demand distinct compute and memory-bandwidth capabilities. As a result, a single homogeneous GPU system now struggles to support agentic inference, motivating an industry shift toward heterogeneous systems with disaggregated serving capabilities, such as the emerging Vera-Rubin pl"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.03741","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/2608.03741/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":"2608.03741","created_at":"2026-08-05T01:37:00.151768+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.03741v1","created_at":"2026-08-05T01:37:00.151768+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.03741","created_at":"2026-08-05T01:37:00.151768+00:00"},{"alias_kind":"pith_short_12","alias_value":"QD3PX54SC5BQ","created_at":"2026-08-05T01:37:00.151768+00:00"},{"alias_kind":"pith_short_16","alias_value":"QD3PX54SC5BQH26N","created_at":"2026-08-05T01:37:00.151768+00:00"},{"alias_kind":"pith_short_8","alias_value":"QD3PX54S","created_at":"2026-08-05T01:37:00.151768+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/QD3PX54SC5BQH26N6YSZY66SH4","json":"https://pith.science/pith/QD3PX54SC5BQH26N6YSZY66SH4.json","graph_json":"https://pith.science/api/pith-number/QD3PX54SC5BQH26N6YSZY66SH4/graph.json","events_json":"https://pith.science/api/pith-number/QD3PX54SC5BQH26N6YSZY66SH4/events.json","paper":"https://pith.science/paper/QD3PX54S"},"agent_actions":{"view_html":"https://pith.science/pith/QD3PX54SC5BQH26N6YSZY66SH4","download_json":"https://pith.science/pith/QD3PX54SC5BQH26N6YSZY66SH4.json","view_paper":"https://pith.science/paper/QD3PX54S","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.03741&json=true","fetch_graph":"https://pith.science/api/pith-number/QD3PX54SC5BQH26N6YSZY66SH4/graph.json","fetch_events":"https://pith.science/api/pith-number/QD3PX54SC5BQH26N6YSZY66SH4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QD3PX54SC5BQH26N6YSZY66SH4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QD3PX54SC5BQH26N6YSZY66SH4/action/storage_attestation","attest_author":"https://pith.science/pith/QD3PX54SC5BQH26N6YSZY66SH4/action/author_attestation","sign_citation":"https://pith.science/pith/QD3PX54SC5BQH26N6YSZY66SH4/action/citation_signature","submit_replication":"https://pith.science/pith/QD3PX54SC5BQH26N6YSZY66SH4/action/replication_record"}},"created_at":"2026-08-05T01:37:00.151768+00:00","updated_at":"2026-08-05T01:37:00.151768+00:00"}