{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:NFAY3UHDOT22WPM6RW76G2CNFL","short_pith_number":"pith:NFAY3UHD","schema_version":"1.0","canonical_sha256":"69418dd0e374f5ab3d9e8dbfe3684d2acf4a8a0ebdc5292fcf7a8eed64d01aab","source":{"kind":"arxiv","id":"2607.08993","version":1},"attestation_state":"computed","paper":{"title":"StreamDQ: Near-Memory Weight DeQuantization in Custom HBM for Scalable AI Inference Acceleration","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AR","authors_text":"Daegun Yoon, Hoshik Kim, Hyeonseok Ju, Ieryung Park, Joonseop Sim, Minki Jeong, Nameun Kang, Seungyong Lee, Soohong Ahn, Youngpyo Joo","submitted_at":"2026-07-09T23:52:57Z","abstract_excerpt":"As large language models (LLMs) scale, their memory and computation demands have grown substantially, making weight-only quantization a widely adopted technique for reducing model size with minimal accuracy loss. However, on current GPUs, CUDA-core-based dequantization introduces substantial instruction overhead, on-chip traffic, and pipeline stalls, making it a major bottleneck for high-throughput, cloud-scale LLM serving. To address these limitations, we propose StreamDQ, a lightweight architectural enhancement that enables on-the-fly dequantization in the memory subsystem for high-throughpu"},"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":"2607.08993","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AR","submitted_at":"2026-07-09T23:52:57Z","cross_cats_sorted":[],"title_canon_sha256":"0912ba169957e106143c0939871b33b362a8ad04e2713d21a82e34716be3470b","abstract_canon_sha256":"9954bd74493cbee4f189f61d088ed18035ad14cf6113d67e1aeca1aab568a174"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-13T00:17:32.775881Z","signature_b64":"zAdK8SAyL6Nj65d1mvwYT7S4NdNhpmJHC6/pbSvc3nJYf52Fz6z28YAX8QKJRxUrxOuLd0oG/c/Dw3M0ox9TDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"69418dd0e374f5ab3d9e8dbfe3684d2acf4a8a0ebdc5292fcf7a8eed64d01aab","last_reissued_at":"2026-07-13T00:17:32.774503Z","signature_status":"signed_v1","first_computed_at":"2026-07-13T00:17:32.774503Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"StreamDQ: Near-Memory Weight DeQuantization in Custom HBM for Scalable AI Inference Acceleration","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AR","authors_text":"Daegun Yoon, Hoshik Kim, Hyeonseok Ju, Ieryung Park, Joonseop Sim, Minki Jeong, Nameun Kang, Seungyong Lee, Soohong Ahn, Youngpyo Joo","submitted_at":"2026-07-09T23:52:57Z","abstract_excerpt":"As large language models (LLMs) scale, their memory and computation demands have grown substantially, making weight-only quantization a widely adopted technique for reducing model size with minimal accuracy loss. However, on current GPUs, CUDA-core-based dequantization introduces substantial instruction overhead, on-chip traffic, and pipeline stalls, making it a major bottleneck for high-throughput, cloud-scale LLM serving. To address these limitations, we propose StreamDQ, a lightweight architectural enhancement that enables on-the-fly dequantization in the memory subsystem for high-throughpu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.08993","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/2607.08993/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":"2607.08993","created_at":"2026-07-13T00:17:32.775201+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.08993v1","created_at":"2026-07-13T00:17:32.775201+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.08993","created_at":"2026-07-13T00:17:32.775201+00:00"},{"alias_kind":"pith_short_12","alias_value":"NFAY3UHDOT22","created_at":"2026-07-13T00:17:32.775201+00:00"},{"alias_kind":"pith_short_16","alias_value":"NFAY3UHDOT22WPM6","created_at":"2026-07-13T00:17:32.775201+00:00"},{"alias_kind":"pith_short_8","alias_value":"NFAY3UHD","created_at":"2026-07-13T00:17:32.775201+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/NFAY3UHDOT22WPM6RW76G2CNFL","json":"https://pith.science/pith/NFAY3UHDOT22WPM6RW76G2CNFL.json","graph_json":"https://pith.science/api/pith-number/NFAY3UHDOT22WPM6RW76G2CNFL/graph.json","events_json":"https://pith.science/api/pith-number/NFAY3UHDOT22WPM6RW76G2CNFL/events.json","paper":"https://pith.science/paper/NFAY3UHD"},"agent_actions":{"view_html":"https://pith.science/pith/NFAY3UHDOT22WPM6RW76G2CNFL","download_json":"https://pith.science/pith/NFAY3UHDOT22WPM6RW76G2CNFL.json","view_paper":"https://pith.science/paper/NFAY3UHD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.08993&json=true","fetch_graph":"https://pith.science/api/pith-number/NFAY3UHDOT22WPM6RW76G2CNFL/graph.json","fetch_events":"https://pith.science/api/pith-number/NFAY3UHDOT22WPM6RW76G2CNFL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NFAY3UHDOT22WPM6RW76G2CNFL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NFAY3UHDOT22WPM6RW76G2CNFL/action/storage_attestation","attest_author":"https://pith.science/pith/NFAY3UHDOT22WPM6RW76G2CNFL/action/author_attestation","sign_citation":"https://pith.science/pith/NFAY3UHDOT22WPM6RW76G2CNFL/action/citation_signature","submit_replication":"https://pith.science/pith/NFAY3UHDOT22WPM6RW76G2CNFL/action/replication_record"}},"created_at":"2026-07-13T00:17:32.775201+00:00","updated_at":"2026-07-13T00:17:32.775201+00:00"}