{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:LZ3SSDTR4M6NH77BM3WYBQBYTD","short_pith_number":"pith:LZ3SSDTR","schema_version":"1.0","canonical_sha256":"5e77290e71e33cd3ffe166ed80c03898e7bcd8652947a390f7c3360f902df762","source":{"kind":"arxiv","id":"2608.12103","version":1},"attestation_state":"computed","paper":{"title":"Who Should Own the Expert Cache? Kernel-Managed Tiering for Trillion-Parameter MoE Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.OS","authors_text":"Daming Li, Jialu Zhang, Yuan Si, Yufeng Lin","submitted_at":"2026-08-12T14:24:09Z","abstract_excerpt":"Mixture-of-experts models whose expert pools dwarf DRAM force every serving system to contain a cache, yet existing systems typically implement this cache in user space using expert-granular, frequency-ranked, explicitly pinned tiers. We characterize the alternative already provided by the OS: the page cache as the expert tier. We use router traces from three MoE models spanning 128 to 896 experts per layer, including a production trillion-parameter model with a 1.45 TB expert pool, and replay them natively against the full pool on GH200 nodes with capacity enforced by three independent mechan"},"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":"2608.12103","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.OS","submitted_at":"2026-08-12T14:24:09Z","cross_cats_sorted":[],"title_canon_sha256":"f96fbb7c277a72e7ef60660674e540d734e0ee1b5d6ddaecc5c091f9a51f9a60","abstract_canon_sha256":"225e478e842e278fee70cebc45b44329f9ae0564075d55716397ffa5862047fe"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-13T01:29:59.566171Z","signature_b64":"UQjLVuBkzaSIjy2cGEwqXQGjB/tmOy7Nvz8mLgR8SWMupoifQdEFISQl8oqNN99iB+WqL3ZByjvGFVWuYIbnDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5e77290e71e33cd3ffe166ed80c03898e7bcd8652947a390f7c3360f902df762","last_reissued_at":"2026-08-13T01:29:59.563506Z","signature_status":"signed_v1","first_computed_at":"2026-08-13T01:29:59.563506Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Who Should Own the Expert Cache? Kernel-Managed Tiering for Trillion-Parameter MoE Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.OS","authors_text":"Daming Li, Jialu Zhang, Yuan Si, Yufeng Lin","submitted_at":"2026-08-12T14:24:09Z","abstract_excerpt":"Mixture-of-experts models whose expert pools dwarf DRAM force every serving system to contain a cache, yet existing systems typically implement this cache in user space using expert-granular, frequency-ranked, explicitly pinned tiers. We characterize the alternative already provided by the OS: the page cache as the expert tier. We use router traces from three MoE models spanning 128 to 896 experts per layer, including a production trillion-parameter model with a 1.45 TB expert pool, and replay them natively against the full pool on GH200 nodes with capacity enforced by three independent mechan"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.12103","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/2608.12103/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":"2608.12103","created_at":"2026-08-13T01:29:59.564954+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.12103v1","created_at":"2026-08-13T01:29:59.564954+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.12103","created_at":"2026-08-13T01:29:59.564954+00:00"},{"alias_kind":"pith_short_12","alias_value":"LZ3SSDTR4M6N","created_at":"2026-08-13T01:29:59.564954+00:00"},{"alias_kind":"pith_short_16","alias_value":"LZ3SSDTR4M6NH77B","created_at":"2026-08-13T01:29:59.564954+00:00"},{"alias_kind":"pith_short_8","alias_value":"LZ3SSDTR","created_at":"2026-08-13T01:29:59.564954+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/LZ3SSDTR4M6NH77BM3WYBQBYTD","json":"https://pith.science/pith/LZ3SSDTR4M6NH77BM3WYBQBYTD.json","graph_json":"https://pith.science/api/pith-number/LZ3SSDTR4M6NH77BM3WYBQBYTD/graph.json","events_json":"https://pith.science/api/pith-number/LZ3SSDTR4M6NH77BM3WYBQBYTD/events.json","paper":"https://pith.science/paper/LZ3SSDTR"},"agent_actions":{"view_html":"https://pith.science/pith/LZ3SSDTR4M6NH77BM3WYBQBYTD","download_json":"https://pith.science/pith/LZ3SSDTR4M6NH77BM3WYBQBYTD.json","view_paper":"https://pith.science/paper/LZ3SSDTR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.12103&json=true","fetch_graph":"https://pith.science/api/pith-number/LZ3SSDTR4M6NH77BM3WYBQBYTD/graph.json","fetch_events":"https://pith.science/api/pith-number/LZ3SSDTR4M6NH77BM3WYBQBYTD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LZ3SSDTR4M6NH77BM3WYBQBYTD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LZ3SSDTR4M6NH77BM3WYBQBYTD/action/storage_attestation","attest_author":"https://pith.science/pith/LZ3SSDTR4M6NH77BM3WYBQBYTD/action/author_attestation","sign_citation":"https://pith.science/pith/LZ3SSDTR4M6NH77BM3WYBQBYTD/action/citation_signature","submit_replication":"https://pith.science/pith/LZ3SSDTR4M6NH77BM3WYBQBYTD/action/replication_record"}},"created_at":"2026-08-13T01:29:59.564954+00:00","updated_at":"2026-08-13T01:29:59.564954+00:00"}