{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:YI53UIENCWYTCTCTR6NBSBK7FU","short_pith_number":"pith:YI53UIEN","schema_version":"1.0","canonical_sha256":"c23bba208d15b1314c538f9a19055f2d1d0685be8fcf2a7e01c19a907ec4c722","source":{"kind":"arxiv","id":"2608.01250","version":1},"attestation_state":"computed","paper":{"title":"Smoothing the Ramp, Not the Peak: Scheduling-Induced Power Dynamics of LLM Inference and Their Grid-Scale Consequences","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.SY"],"primary_cat":"eess.SY","authors_text":"Dai Wang, Pan Li, Xia Miao, Yize Chen","submitted_at":"2026-08-02T14:01:51Z","abstract_excerpt":"Large language model (LLM) inference serving is a fast-growing electricity load whose power dynamics remain uncharacterized from a grid-planning perspective. Using real, measured GPU power traces, we show that chunked prefill scheduling, a latency-motivated technique already deployed by default in production LLM serving, is a controllable knob that regulates power ramp rate without touching peak power. Contrary to the intuitive hypothesis that splitting a long prompt's computation into smaller steps should flatten its power spike, peak power stays relatively the same while mean ramp rate falls"},"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":"2608.01250","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"eess.SY","submitted_at":"2026-08-02T14:01:51Z","cross_cats_sorted":["cs.SY"],"title_canon_sha256":"e3856bf71d5ddba623e0d11f4aff67ce0dcf58d19d3e408ea6e2108597d49e4a","abstract_canon_sha256":"82ada1240f8bdffc1cfaeb5a6ae8edd99441657130ccb2f60d832a53bdc07f07"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-04T02:02:16.820091Z","signature_b64":"FehpsnsvWHBl1cq8KV2tfSZFQCfduvBtoXcVAcoTtaQEc7tZIRRVumLJEXYC9Tcuy8XKd3OHmgm4Ux5WLyybAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c23bba208d15b1314c538f9a19055f2d1d0685be8fcf2a7e01c19a907ec4c722","last_reissued_at":"2026-08-04T02:02:16.818475Z","signature_status":"signed_v1","first_computed_at":"2026-08-04T02:02:16.818475Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Smoothing the Ramp, Not the Peak: Scheduling-Induced Power Dynamics of LLM Inference and Their Grid-Scale Consequences","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.SY"],"primary_cat":"eess.SY","authors_text":"Dai Wang, Pan Li, Xia Miao, Yize Chen","submitted_at":"2026-08-02T14:01:51Z","abstract_excerpt":"Large language model (LLM) inference serving is a fast-growing electricity load whose power dynamics remain uncharacterized from a grid-planning perspective. Using real, measured GPU power traces, we show that chunked prefill scheduling, a latency-motivated technique already deployed by default in production LLM serving, is a controllable knob that regulates power ramp rate without touching peak power. Contrary to the intuitive hypothesis that splitting a long prompt's computation into smaller steps should flatten its power spike, peak power stays relatively the same while mean ramp rate falls"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.01250","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/2608.01250/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":"2608.01250","created_at":"2026-08-04T02:02:16.819933+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.01250v1","created_at":"2026-08-04T02:02:16.819933+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.01250","created_at":"2026-08-04T02:02:16.819933+00:00"},{"alias_kind":"pith_short_12","alias_value":"YI53UIENCWYT","created_at":"2026-08-04T02:02:16.819933+00:00"},{"alias_kind":"pith_short_16","alias_value":"YI53UIENCWYTCTCT","created_at":"2026-08-04T02:02:16.819933+00:00"},{"alias_kind":"pith_short_8","alias_value":"YI53UIEN","created_at":"2026-08-04T02:02:16.819933+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/YI53UIENCWYTCTCTR6NBSBK7FU","json":"https://pith.science/pith/YI53UIENCWYTCTCTR6NBSBK7FU.json","graph_json":"https://pith.science/api/pith-number/YI53UIENCWYTCTCTR6NBSBK7FU/graph.json","events_json":"https://pith.science/api/pith-number/YI53UIENCWYTCTCTR6NBSBK7FU/events.json","paper":"https://pith.science/paper/YI53UIEN"},"agent_actions":{"view_html":"https://pith.science/pith/YI53UIENCWYTCTCTR6NBSBK7FU","download_json":"https://pith.science/pith/YI53UIENCWYTCTCTR6NBSBK7FU.json","view_paper":"https://pith.science/paper/YI53UIEN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.01250&json=true","fetch_graph":"https://pith.science/api/pith-number/YI53UIENCWYTCTCTR6NBSBK7FU/graph.json","fetch_events":"https://pith.science/api/pith-number/YI53UIENCWYTCTCTR6NBSBK7FU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YI53UIENCWYTCTCTR6NBSBK7FU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YI53UIENCWYTCTCTR6NBSBK7FU/action/storage_attestation","attest_author":"https://pith.science/pith/YI53UIENCWYTCTCTR6NBSBK7FU/action/author_attestation","sign_citation":"https://pith.science/pith/YI53UIENCWYTCTCTR6NBSBK7FU/action/citation_signature","submit_replication":"https://pith.science/pith/YI53UIENCWYTCTCTR6NBSBK7FU/action/replication_record"}},"created_at":"2026-08-04T02:02:16.819933+00:00","updated_at":"2026-08-04T02:02:16.819933+00:00"}