{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:NPZBV2NCMOQAVBF7KFVTGXAGUV","short_pith_number":"pith:NPZBV2NC","schema_version":"1.0","canonical_sha256":"6bf21ae9a263a00a84bf516b335c06a54c85ab88cd4f7bf157e67f5a2017a91c","source":{"kind":"arxiv","id":"2507.11067","version":1},"attestation_state":"computed","paper":{"title":"MMStencil: Optimizing High-order Stencils on Multicore CPU using Matrix Unit","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.DC","authors_text":"Guangwen Yang, Jiayu Fu, Lanke He, Lin Gan, Tianqi Mao, Wei Xue, Wenqiang Wang, Wubing Wan, Yinuo Wang, Zekun Yin, Zeyu Song","submitted_at":"2025-07-15T08:00:11Z","abstract_excerpt":"Matrix-accelerated stencil computation is a hot research topic, yet its application to three-dimensional (3D) high-order stencils and HPC remains underexplored. With the emergence of matrix units on multicore CPUs, we analyze matrix-based acceleration strategies and tailor an optimal approach for 3D high-order stencils. We introduce algorithmic optimizations based on SIMD and matrix units to address strided memory accesses, alignment conflicts, and redundant accesses. We propose memory optimizations to boost on-package memory efficiency, and a novel multi-thread parallelism paradigm to overcom"},"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":"2507.11067","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.DC","submitted_at":"2025-07-15T08:00:11Z","cross_cats_sorted":[],"title_canon_sha256":"b66c6fb6fc4e88e627315b12b91e450cb94739f01e60a880a03371fe5cd74dfe","abstract_canon_sha256":"bc3a9b94501a76f7f2592d6765a38f207d150fbcc67fda09993d1eb7cfa70a5f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:37:31.351218Z","signature_b64":"7QafEU0Tp5QIRqYm6u4LQqRehOQDZQQ2mrwOCQEaYXjuVh2mzQsBe8zmULvRDXLYYOs5ucBDy3Ivg4etOk6/Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6bf21ae9a263a00a84bf516b335c06a54c85ab88cd4f7bf157e67f5a2017a91c","last_reissued_at":"2026-07-05T11:37:31.350710Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:37:31.350710Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MMStencil: Optimizing High-order Stencils on Multicore CPU using Matrix Unit","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.DC","authors_text":"Guangwen Yang, Jiayu Fu, Lanke He, Lin Gan, Tianqi Mao, Wei Xue, Wenqiang Wang, Wubing Wan, Yinuo Wang, Zekun Yin, Zeyu Song","submitted_at":"2025-07-15T08:00:11Z","abstract_excerpt":"Matrix-accelerated stencil computation is a hot research topic, yet its application to three-dimensional (3D) high-order stencils and HPC remains underexplored. With the emergence of matrix units on multicore CPUs, we analyze matrix-based acceleration strategies and tailor an optimal approach for 3D high-order stencils. We introduce algorithmic optimizations based on SIMD and matrix units to address strided memory accesses, alignment conflicts, and redundant accesses. We propose memory optimizations to boost on-package memory efficiency, and a novel multi-thread parallelism paradigm to overcom"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.11067","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/2507.11067/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":"2507.11067","created_at":"2026-07-05T11:37:31.350777+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.11067v1","created_at":"2026-07-05T11:37:31.350777+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.11067","created_at":"2026-07-05T11:37:31.350777+00:00"},{"alias_kind":"pith_short_12","alias_value":"NPZBV2NCMOQA","created_at":"2026-07-05T11:37:31.350777+00:00"},{"alias_kind":"pith_short_16","alias_value":"NPZBV2NCMOQAVBF7","created_at":"2026-07-05T11:37:31.350777+00:00"},{"alias_kind":"pith_short_8","alias_value":"NPZBV2NC","created_at":"2026-07-05T11:37:31.350777+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/NPZBV2NCMOQAVBF7KFVTGXAGUV","json":"https://pith.science/pith/NPZBV2NCMOQAVBF7KFVTGXAGUV.json","graph_json":"https://pith.science/api/pith-number/NPZBV2NCMOQAVBF7KFVTGXAGUV/graph.json","events_json":"https://pith.science/api/pith-number/NPZBV2NCMOQAVBF7KFVTGXAGUV/events.json","paper":"https://pith.science/paper/NPZBV2NC"},"agent_actions":{"view_html":"https://pith.science/pith/NPZBV2NCMOQAVBF7KFVTGXAGUV","download_json":"https://pith.science/pith/NPZBV2NCMOQAVBF7KFVTGXAGUV.json","view_paper":"https://pith.science/paper/NPZBV2NC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.11067&json=true","fetch_graph":"https://pith.science/api/pith-number/NPZBV2NCMOQAVBF7KFVTGXAGUV/graph.json","fetch_events":"https://pith.science/api/pith-number/NPZBV2NCMOQAVBF7KFVTGXAGUV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NPZBV2NCMOQAVBF7KFVTGXAGUV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NPZBV2NCMOQAVBF7KFVTGXAGUV/action/storage_attestation","attest_author":"https://pith.science/pith/NPZBV2NCMOQAVBF7KFVTGXAGUV/action/author_attestation","sign_citation":"https://pith.science/pith/NPZBV2NCMOQAVBF7KFVTGXAGUV/action/citation_signature","submit_replication":"https://pith.science/pith/NPZBV2NCMOQAVBF7KFVTGXAGUV/action/replication_record"}},"created_at":"2026-07-05T11:37:31.350777+00:00","updated_at":"2026-07-05T11:37:31.350777+00:00"}