{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:RUBQMX3IU3CWRQXCML63Q7IGMM","short_pith_number":"pith:RUBQMX3I","schema_version":"1.0","canonical_sha256":"8d03065f68a6c568c2e262fdb87d0663130cd7fab05138bc5d766638306d57e5","source":{"kind":"arxiv","id":"2605.17889","version":1},"attestation_state":"computed","paper":{"title":"CoX-MoE: Coalesced Expert Execution for High-Throughput MoE Inference with AMX-Enabled CPU-GPU Co-Execution","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Mu-Young Son, Seungjae Yoo, Soongyu Choi Joo-Young Kim, Yi Chen","submitted_at":"2026-05-18T05:54:30Z","abstract_excerpt":"The Mixture-of-Experts (MoE) architecture improves computational efficiency via sparse expert activation, but throughput-oriented inference faces substantial GPU memory pressure due to a significant parameter size and intermediate data. Prior works attempt to mitigate this using expert offloading with micro-batching or by offloading computation to the CPU. However, the fragmented workload resulting from micro-batching degrades operational intensity, causing expert execution to become memory-bound. Meanwhile, CPU offloading is constrained by slow PCIe transfers and its limited applicability to "},"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":"2605.17889","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-18T05:54:30Z","cross_cats_sorted":[],"title_canon_sha256":"26520978c289f360719fcc43ef5e56eb3dd1d6109f95d835b56a9b028824f3d9","abstract_canon_sha256":"27876153aece4090016952a83f9c2d196ce2835b13642d3b0a0542702147f260"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:05:03.944906Z","signature_b64":"6BcaedpZzIi38we9Yo4IpXk6PKAuSwQP6VCK0vG0ZjUJpnr/v6tm10KMzayZPHaM+f2Qwywo+u9vIJsHmp1MDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8d03065f68a6c568c2e262fdb87d0663130cd7fab05138bc5d766638306d57e5","last_reissued_at":"2026-05-20T00:05:03.944014Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:05:03.944014Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CoX-MoE: Coalesced Expert Execution for High-Throughput MoE Inference with AMX-Enabled CPU-GPU Co-Execution","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Mu-Young Son, Seungjae Yoo, Soongyu Choi Joo-Young Kim, Yi Chen","submitted_at":"2026-05-18T05:54:30Z","abstract_excerpt":"The Mixture-of-Experts (MoE) architecture improves computational efficiency via sparse expert activation, but throughput-oriented inference faces substantial GPU memory pressure due to a significant parameter size and intermediate data. Prior works attempt to mitigate this using expert offloading with micro-batching or by offloading computation to the CPU. However, the fragmented workload resulting from micro-batching degrades operational intensity, causing expert execution to become memory-bound. Meanwhile, CPU offloading is constrained by slow PCIe transfers and its limited applicability to "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.17889","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/2605.17889/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":"2605.17889","created_at":"2026-05-20T00:05:03.944167+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.17889v1","created_at":"2026-05-20T00:05:03.944167+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.17889","created_at":"2026-05-20T00:05:03.944167+00:00"},{"alias_kind":"pith_short_12","alias_value":"RUBQMX3IU3CW","created_at":"2026-05-20T00:05:03.944167+00:00"},{"alias_kind":"pith_short_16","alias_value":"RUBQMX3IU3CWRQXC","created_at":"2026-05-20T00:05:03.944167+00:00"},{"alias_kind":"pith_short_8","alias_value":"RUBQMX3I","created_at":"2026-05-20T00:05:03.944167+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/RUBQMX3IU3CWRQXCML63Q7IGMM","json":"https://pith.science/pith/RUBQMX3IU3CWRQXCML63Q7IGMM.json","graph_json":"https://pith.science/api/pith-number/RUBQMX3IU3CWRQXCML63Q7IGMM/graph.json","events_json":"https://pith.science/api/pith-number/RUBQMX3IU3CWRQXCML63Q7IGMM/events.json","paper":"https://pith.science/paper/RUBQMX3I"},"agent_actions":{"view_html":"https://pith.science/pith/RUBQMX3IU3CWRQXCML63Q7IGMM","download_json":"https://pith.science/pith/RUBQMX3IU3CWRQXCML63Q7IGMM.json","view_paper":"https://pith.science/paper/RUBQMX3I","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.17889&json=true","fetch_graph":"https://pith.science/api/pith-number/RUBQMX3IU3CWRQXCML63Q7IGMM/graph.json","fetch_events":"https://pith.science/api/pith-number/RUBQMX3IU3CWRQXCML63Q7IGMM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RUBQMX3IU3CWRQXCML63Q7IGMM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RUBQMX3IU3CWRQXCML63Q7IGMM/action/storage_attestation","attest_author":"https://pith.science/pith/RUBQMX3IU3CWRQXCML63Q7IGMM/action/author_attestation","sign_citation":"https://pith.science/pith/RUBQMX3IU3CWRQXCML63Q7IGMM/action/citation_signature","submit_replication":"https://pith.science/pith/RUBQMX3IU3CWRQXCML63Q7IGMM/action/replication_record"}},"created_at":"2026-05-20T00:05:03.944167+00:00","updated_at":"2026-05-20T00:05:03.944167+00:00"}