{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:VGRENWABU3LASKCH76WFDAOOX4","short_pith_number":"pith:VGRENWAB","schema_version":"1.0","canonical_sha256":"a9a246d801a6d6092847ffac5181cebf1a68cda2e751789079643b37fde65d78","source":{"kind":"arxiv","id":"2606.25353","version":1},"attestation_state":"computed","paper":{"title":"Cache-Resident LLM Inference in GB-Scale Last-Level Caches","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AR","authors_text":"Ceyu Xu, Jian Weng, Marco Canini, Tongzhou Gu, Wanning Zhang","submitted_at":"2026-06-24T03:41:30Z","abstract_excerpt":"Large language model (LLM) inference is increasingly dominated by data movement across the memory hierarchy. Recent 3D-stacked cache technologies have enabled GB-scale last-level caches in modern server CPUs, making it possible to keep reusable model weights on chip and exploit cache bandwidth and latency. Achieving this regime is not straightforward: deeper pipelining for weight residency increases in-flight requests and KV-cache footprint, while cache-resident operators make operator-boundary synchronization a visible bottleneck.\n  We present a cache-resident execution model for inference on"},"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":"2606.25353","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AR","submitted_at":"2026-06-24T03:41:30Z","cross_cats_sorted":[],"title_canon_sha256":"1749ebd7eceea12fff89bae9e0b934a5187437ffde4f5c798ac1df37062a7988","abstract_canon_sha256":"9315c1d305c15d7ab8a9fe7da8d9d530f01f8228e217b666a72a1793b381b7cb"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-25T01:18:02.898955Z","signature_b64":"jpdstxlsauJ6ZO6v511M7QfDLp4o1z1NHapglCqWJAWW5ZwDFr78JDDxW9d7TplOV+9waAOdhuQbG8q9JZJmCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a9a246d801a6d6092847ffac5181cebf1a68cda2e751789079643b37fde65d78","last_reissued_at":"2026-06-25T01:18:02.898591Z","signature_status":"signed_v1","first_computed_at":"2026-06-25T01:18:02.898591Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Cache-Resident LLM Inference in GB-Scale Last-Level Caches","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AR","authors_text":"Ceyu Xu, Jian Weng, Marco Canini, Tongzhou Gu, Wanning Zhang","submitted_at":"2026-06-24T03:41:30Z","abstract_excerpt":"Large language model (LLM) inference is increasingly dominated by data movement across the memory hierarchy. Recent 3D-stacked cache technologies have enabled GB-scale last-level caches in modern server CPUs, making it possible to keep reusable model weights on chip and exploit cache bandwidth and latency. Achieving this regime is not straightforward: deeper pipelining for weight residency increases in-flight requests and KV-cache footprint, while cache-resident operators make operator-boundary synchronization a visible bottleneck.\n  We present a cache-resident execution model for inference on"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.25353","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/2606.25353/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":"2606.25353","created_at":"2026-06-25T01:18:02.898663+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.25353v1","created_at":"2026-06-25T01:18:02.898663+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.25353","created_at":"2026-06-25T01:18:02.898663+00:00"},{"alias_kind":"pith_short_12","alias_value":"VGRENWABU3LA","created_at":"2026-06-25T01:18:02.898663+00:00"},{"alias_kind":"pith_short_16","alias_value":"VGRENWABU3LASKCH","created_at":"2026-06-25T01:18:02.898663+00:00"},{"alias_kind":"pith_short_8","alias_value":"VGRENWAB","created_at":"2026-06-25T01:18:02.898663+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/VGRENWABU3LASKCH76WFDAOOX4","json":"https://pith.science/pith/VGRENWABU3LASKCH76WFDAOOX4.json","graph_json":"https://pith.science/api/pith-number/VGRENWABU3LASKCH76WFDAOOX4/graph.json","events_json":"https://pith.science/api/pith-number/VGRENWABU3LASKCH76WFDAOOX4/events.json","paper":"https://pith.science/paper/VGRENWAB"},"agent_actions":{"view_html":"https://pith.science/pith/VGRENWABU3LASKCH76WFDAOOX4","download_json":"https://pith.science/pith/VGRENWABU3LASKCH76WFDAOOX4.json","view_paper":"https://pith.science/paper/VGRENWAB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.25353&json=true","fetch_graph":"https://pith.science/api/pith-number/VGRENWABU3LASKCH76WFDAOOX4/graph.json","fetch_events":"https://pith.science/api/pith-number/VGRENWABU3LASKCH76WFDAOOX4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VGRENWABU3LASKCH76WFDAOOX4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VGRENWABU3LASKCH76WFDAOOX4/action/storage_attestation","attest_author":"https://pith.science/pith/VGRENWABU3LASKCH76WFDAOOX4/action/author_attestation","sign_citation":"https://pith.science/pith/VGRENWABU3LASKCH76WFDAOOX4/action/citation_signature","submit_replication":"https://pith.science/pith/VGRENWABU3LASKCH76WFDAOOX4/action/replication_record"}},"created_at":"2026-06-25T01:18:02.898663+00:00","updated_at":"2026-06-25T01:18:02.898663+00:00"}