{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:UV5ZLYXKRTCCL43BYDHMIR5PRD","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"829a1182ae42f40d5963eea18d3e9d979707672237f42e8e4ccc44a1ccbb86e3","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2025-03-11T11:21:35Z","title_canon_sha256":"584d60c9bed26f9d807aeb0e3d917ca7419fcad35d7b83d8e7cb8813f195febd"},"schema_version":"1.0","source":{"id":"2503.08311","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.08311","created_at":"2026-07-05T11:35:33Z"},{"alias_kind":"arxiv_version","alias_value":"2503.08311v2","created_at":"2026-07-05T11:35:33Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.08311","created_at":"2026-07-05T11:35:33Z"},{"alias_kind":"pith_short_12","alias_value":"UV5ZLYXKRTCC","created_at":"2026-07-05T11:35:33Z"},{"alias_kind":"pith_short_16","alias_value":"UV5ZLYXKRTCCL43B","created_at":"2026-07-05T11:35:33Z"},{"alias_kind":"pith_short_8","alias_value":"UV5ZLYXK","created_at":"2026-07-05T11:35:33Z"}],"graph_snapshots":[{"event_id":"sha256:8713d5ef838be2c08fb8980e9c722a6488aec432e813b1548b52a11cf707f53a","target":"graph","created_at":"2026-07-05T11:35:33Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2503.08311/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models have been widely adopted across different tasks, but their auto-regressive generation nature often leads to inefficient resource utilization during inference. While batching is commonly used to increase throughput, performance gains plateau beyond a certain batch size, especially with smaller models, a phenomenon that existing literature typically explains as a shift to the compute-bound regime. In this paper, through an in-depth GPU-level analysis, we reveal that large-batch inference remains memory-bound, with most GPU compute capabilities underutilized due to DRAM band","authors_text":"Chen Wang, Eun Kyung Lee, Ferran Agullo, Jordi Torres, Josep Ll. Berral, Olivier Tardieu, Pol G. Recasens, Yue Zhu","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2025-03-11T11:21:35Z","title":"Mind the Memory Gap: Unveiling GPU Bottlenecks in Large-Batch LLM Inference"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.08311","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:21146c3a027d3d77449943b5ffab997d4c8a6872bad347be5a64611b26158833","target":"record","created_at":"2026-07-05T11:35:33Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"829a1182ae42f40d5963eea18d3e9d979707672237f42e8e4ccc44a1ccbb86e3","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.DC","submitted_at":"2025-03-11T11:21:35Z","title_canon_sha256":"584d60c9bed26f9d807aeb0e3d917ca7419fcad35d7b83d8e7cb8813f195febd"},"schema_version":"1.0","source":{"id":"2503.08311","kind":"arxiv","version":2}},"canonical_sha256":"a57b95e2ea8cc425f361c0cec447af88d0533e8ac78d74112bf53981fe507ad7","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a57b95e2ea8cc425f361c0cec447af88d0533e8ac78d74112bf53981fe507ad7","first_computed_at":"2026-07-05T11:35:33.827340Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:35:33.827340Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"HBtZXECmBeRiBoIEB8wROPaCvmvPdNTBLfnikZIP7kBI9cinWxskz7H2LdlTk4UpOlLr8aggfchXswywCve3BA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:35:33.827772Z","signed_message":"canonical_sha256_bytes"},"source_id":"2503.08311","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:21146c3a027d3d77449943b5ffab997d4c8a6872bad347be5a64611b26158833","sha256:8713d5ef838be2c08fb8980e9c722a6488aec432e813b1548b52a11cf707f53a"],"state_sha256":"75431fbe93c7a61ef9cc0a915caf76a72c95d09b2f1e047073c781d36265228e"}