{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:LOE26RDX22GQIUSZ5DNNP2PJ2L","short_pith_number":"pith:LOE26RDX","schema_version":"1.0","canonical_sha256":"5b89af4477d68d045259e8dad7e9e9d2ff244424dc4a1b88fc48995bd2ecdc36","source":{"kind":"arxiv","id":"2502.04563","version":3},"attestation_state":"computed","paper":{"title":"WaferLLM: Large Language Model Inference at Wafer Scale","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.AR","cs.DC","cs.ET"],"primary_cat":"cs.LG","authors_text":"Congjie He, Fan Yang, Jilong Xue, Lingxiao Ma, Luo Mai, Pei Mu, Yeqi Huang, Ziming Miao","submitted_at":"2025-02-06T23:32:19Z","abstract_excerpt":"Emerging AI accelerators increasingly adopt wafer-scale manufacturing technologies, integrating hundreds of thousands of AI cores in a mesh architecture with large distributed on-chip memory (tens of GB in total) and ultra-high on-chip memory bandwidth (tens of PB/s). However, current LLM inference systems, optimized for shared memory architectures like GPUs, fail to exploit these accelerators fully.\n  We introduce WaferLLM, the first wafer-scale LLM inference system. WaferLLM is guided by a novel PLMR model (pronounced as \"Plummer\") that captures the unique hardware characteristics of wafer-s"},"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":"2502.04563","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-06T23:32:19Z","cross_cats_sorted":["cs.AI","cs.AR","cs.DC","cs.ET"],"title_canon_sha256":"74f4a4f25156e1cd1e68ad68c6046b449e008bb1279e8e99d3a4860085b4f521","abstract_canon_sha256":"a14d639a52cc24d44dd2716abd3f7851168acaa8a875bd66e9d8e9bc7091bdb9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:12:25.647362Z","signature_b64":"JEziedqD5QQvM4f9B+UrSMkC+gsdhxW0dcqY0/iiaTWdUCsrjnYY6Y6ZhF0GNvfq1FTcwdbbAvt6zp77idu4Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5b89af4477d68d045259e8dad7e9e9d2ff244424dc4a1b88fc48995bd2ecdc36","last_reissued_at":"2026-07-05T11:12:25.646835Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:12:25.646835Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"WaferLLM: Large Language Model Inference at Wafer Scale","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.AR","cs.DC","cs.ET"],"primary_cat":"cs.LG","authors_text":"Congjie He, Fan Yang, Jilong Xue, Lingxiao Ma, Luo Mai, Pei Mu, Yeqi Huang, Ziming Miao","submitted_at":"2025-02-06T23:32:19Z","abstract_excerpt":"Emerging AI accelerators increasingly adopt wafer-scale manufacturing technologies, integrating hundreds of thousands of AI cores in a mesh architecture with large distributed on-chip memory (tens of GB in total) and ultra-high on-chip memory bandwidth (tens of PB/s). However, current LLM inference systems, optimized for shared memory architectures like GPUs, fail to exploit these accelerators fully.\n  We introduce WaferLLM, the first wafer-scale LLM inference system. WaferLLM is guided by a novel PLMR model (pronounced as \"Plummer\") that captures the unique hardware characteristics of wafer-s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.04563","kind":"arxiv","version":3},"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/2502.04563/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":"2502.04563","created_at":"2026-07-05T11:12:25.646899+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.04563v3","created_at":"2026-07-05T11:12:25.646899+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.04563","created_at":"2026-07-05T11:12:25.646899+00:00"},{"alias_kind":"pith_short_12","alias_value":"LOE26RDX22GQ","created_at":"2026-07-05T11:12:25.646899+00:00"},{"alias_kind":"pith_short_16","alias_value":"LOE26RDX22GQIUSZ","created_at":"2026-07-05T11:12:25.646899+00:00"},{"alias_kind":"pith_short_8","alias_value":"LOE26RDX","created_at":"2026-07-05T11:12:25.646899+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.22968","citing_title":"MOCAP: Wafer-Scale-Chip-Oriented Memory-Orchestrated Chunked Pipelining Framework for Prefill-Only LLM Inference","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2606.28754","citing_title":"SHIFT: Dynamic Compute Relocation Framework for Communication-Aware Chiplet-Based Systems","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24203","citing_title":"Agentic Witnessing: Pragmatic and Scalable TEE-Enabled Privacy-Preserving Auditing","ref_index":15,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LOE26RDX22GQIUSZ5DNNP2PJ2L","json":"https://pith.science/pith/LOE26RDX22GQIUSZ5DNNP2PJ2L.json","graph_json":"https://pith.science/api/pith-number/LOE26RDX22GQIUSZ5DNNP2PJ2L/graph.json","events_json":"https://pith.science/api/pith-number/LOE26RDX22GQIUSZ5DNNP2PJ2L/events.json","paper":"https://pith.science/paper/LOE26RDX"},"agent_actions":{"view_html":"https://pith.science/pith/LOE26RDX22GQIUSZ5DNNP2PJ2L","download_json":"https://pith.science/pith/LOE26RDX22GQIUSZ5DNNP2PJ2L.json","view_paper":"https://pith.science/paper/LOE26RDX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.04563&json=true","fetch_graph":"https://pith.science/api/pith-number/LOE26RDX22GQIUSZ5DNNP2PJ2L/graph.json","fetch_events":"https://pith.science/api/pith-number/LOE26RDX22GQIUSZ5DNNP2PJ2L/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LOE26RDX22GQIUSZ5DNNP2PJ2L/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LOE26RDX22GQIUSZ5DNNP2PJ2L/action/storage_attestation","attest_author":"https://pith.science/pith/LOE26RDX22GQIUSZ5DNNP2PJ2L/action/author_attestation","sign_citation":"https://pith.science/pith/LOE26RDX22GQIUSZ5DNNP2PJ2L/action/citation_signature","submit_replication":"https://pith.science/pith/LOE26RDX22GQIUSZ5DNNP2PJ2L/action/replication_record"}},"created_at":"2026-07-05T11:12:25.646899+00:00","updated_at":"2026-07-05T11:12:25.646899+00:00"}