{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:IOSTOJNQH2YWQ5W4735E2LW7G3","short_pith_number":"pith:IOSTOJNQ","schema_version":"1.0","canonical_sha256":"43a53725b03eb16876dcfefa4d2edf36d1309e2535586848d3e5cd796131e11f","source":{"kind":"arxiv","id":"2509.12993","version":3},"attestation_state":"computed","paper":{"title":"HPIM: Heterogeneous Processing-In-Memory-based Accelerator for Large Language Models Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AR","authors_text":"Ao Zhou, Cenlin Duan, Jianlei Yang, Lingkun Long, Rubing Yang, Weisheng Zhao, Xiaolin He, Xueyan Wang, Yikun Wang, Yingjie Qi, Yiou Wang","submitted_at":"2025-09-16T12:04:00Z","abstract_excerpt":"The deployment of large language models (LLMs) presents significant challenges due to their enormous memory footprints, low arithmetic intensity, and stringent latency requirements, particularly during the autoregressive decoding stage. Traditional compute-centric accelerators, such as GPUs, suffer from severe resource underutilization and memory bandwidth bottlenecks in these memory-bound workloads. To overcome these fundamental limitations, we propose HPIM, the first memory-centric heterogeneous Processing-In-Memory (PIM) accelerator that integrates SRAM-PIM and HBM-PIM subsystems designed 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":"2509.12993","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AR","submitted_at":"2025-09-16T12:04:00Z","cross_cats_sorted":[],"title_canon_sha256":"10a068f86cc87f637fa4449b9a0819cc2139ed84c531839c5cbdd8761e2af309","abstract_canon_sha256":"be2ca5caea4cc85fd67f7a9ff9d1acaa35f1ea1c65e6f1db2eba723315ff3742"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-07T01:15:57.911362Z","signature_b64":"nVME5X5GF6NlE7GZ7IHsYiir2nAD9yDXtzvU+EtOQOheZB+m6vJhOui9UoSdesp3I65UdBB1zXLDjWb3+JfBDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"43a53725b03eb16876dcfefa4d2edf36d1309e2535586848d3e5cd796131e11f","last_reissued_at":"2026-07-07T01:15:57.910409Z","signature_status":"signed_v1","first_computed_at":"2026-07-07T01:15:57.910409Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"HPIM: Heterogeneous Processing-In-Memory-based Accelerator for Large Language Models Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AR","authors_text":"Ao Zhou, Cenlin Duan, Jianlei Yang, Lingkun Long, Rubing Yang, Weisheng Zhao, Xiaolin He, Xueyan Wang, Yikun Wang, Yingjie Qi, Yiou Wang","submitted_at":"2025-09-16T12:04:00Z","abstract_excerpt":"The deployment of large language models (LLMs) presents significant challenges due to their enormous memory footprints, low arithmetic intensity, and stringent latency requirements, particularly during the autoregressive decoding stage. Traditional compute-centric accelerators, such as GPUs, suffer from severe resource underutilization and memory bandwidth bottlenecks in these memory-bound workloads. To overcome these fundamental limitations, we propose HPIM, the first memory-centric heterogeneous Processing-In-Memory (PIM) accelerator that integrates SRAM-PIM and HBM-PIM subsystems designed s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.12993","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/2509.12993/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":"2509.12993","created_at":"2026-07-07T01:15:57.910524+00:00"},{"alias_kind":"arxiv_version","alias_value":"2509.12993v3","created_at":"2026-07-07T01:15:57.910524+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.12993","created_at":"2026-07-07T01:15:57.910524+00:00"},{"alias_kind":"pith_short_12","alias_value":"IOSTOJNQH2YW","created_at":"2026-07-07T01:15:57.910524+00:00"},{"alias_kind":"pith_short_16","alias_value":"IOSTOJNQH2YWQ5W4","created_at":"2026-07-07T01:15:57.910524+00:00"},{"alias_kind":"pith_short_8","alias_value":"IOSTOJNQ","created_at":"2026-07-07T01:15:57.910524+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/IOSTOJNQH2YWQ5W4735E2LW7G3","json":"https://pith.science/pith/IOSTOJNQH2YWQ5W4735E2LW7G3.json","graph_json":"https://pith.science/api/pith-number/IOSTOJNQH2YWQ5W4735E2LW7G3/graph.json","events_json":"https://pith.science/api/pith-number/IOSTOJNQH2YWQ5W4735E2LW7G3/events.json","paper":"https://pith.science/paper/IOSTOJNQ"},"agent_actions":{"view_html":"https://pith.science/pith/IOSTOJNQH2YWQ5W4735E2LW7G3","download_json":"https://pith.science/pith/IOSTOJNQH2YWQ5W4735E2LW7G3.json","view_paper":"https://pith.science/paper/IOSTOJNQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2509.12993&json=true","fetch_graph":"https://pith.science/api/pith-number/IOSTOJNQH2YWQ5W4735E2LW7G3/graph.json","fetch_events":"https://pith.science/api/pith-number/IOSTOJNQH2YWQ5W4735E2LW7G3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IOSTOJNQH2YWQ5W4735E2LW7G3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IOSTOJNQH2YWQ5W4735E2LW7G3/action/storage_attestation","attest_author":"https://pith.science/pith/IOSTOJNQH2YWQ5W4735E2LW7G3/action/author_attestation","sign_citation":"https://pith.science/pith/IOSTOJNQH2YWQ5W4735E2LW7G3/action/citation_signature","submit_replication":"https://pith.science/pith/IOSTOJNQH2YWQ5W4735E2LW7G3/action/replication_record"}},"created_at":"2026-07-07T01:15:57.910524+00:00","updated_at":"2026-07-07T01:15:57.910524+00:00"}