{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2022:JOS4IE6KUZH55WH7MIHERJ3ZL7","short_pith_number":"pith:JOS4IE6K","canonical_record":{"source":{"id":"2208.05321","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2022-08-08T12:08:05Z","cross_cats_sorted":["cs.AI","cs.DC","cs.LG"],"title_canon_sha256":"ab94b7c4dd319074cb56a1f23304279c85824fcae557c2ff7d6f9eb70d27d448","abstract_canon_sha256":"92ac121773f811296471d63a2196530d6fc4fe1daaffce11d0cb15910c476efb"},"schema_version":"1.0"},"canonical_sha256":"4ba5c413caa64fded8ff620e48a7795fed8ad4a51e14fe49c524766f7e6ccb9e","source":{"kind":"arxiv","id":"2208.05321","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2208.05321","created_at":"2026-07-05T04:47:34Z"},{"alias_kind":"arxiv_version","alias_value":"2208.05321v1","created_at":"2026-07-05T04:47:34Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2208.05321","created_at":"2026-07-05T04:47:34Z"},{"alias_kind":"pith_short_12","alias_value":"JOS4IE6KUZH5","created_at":"2026-07-05T04:47:34Z"},{"alias_kind":"pith_short_16","alias_value":"JOS4IE6KUZH55WH7","created_at":"2026-07-05T04:47:34Z"},{"alias_kind":"pith_short_8","alias_value":"JOS4IE6K","created_at":"2026-07-05T04:47:34Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2022:JOS4IE6KUZH55WH7MIHERJ3ZL7","target":"record","payload":{"canonical_record":{"source":{"id":"2208.05321","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2022-08-08T12:08:05Z","cross_cats_sorted":["cs.AI","cs.DC","cs.LG"],"title_canon_sha256":"ab94b7c4dd319074cb56a1f23304279c85824fcae557c2ff7d6f9eb70d27d448","abstract_canon_sha256":"92ac121773f811296471d63a2196530d6fc4fe1daaffce11d0cb15910c476efb"},"schema_version":"1.0"},"canonical_sha256":"4ba5c413caa64fded8ff620e48a7795fed8ad4a51e14fe49c524766f7e6ccb9e","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:47:34.822451Z","signature_b64":"hQGv2MLIIdHLQed5Np43jlyIsn46NdY9kicNxvpWRHB/5lASugYsbVrkRInN6ME/0Fg+z7uAL16yKEOb43W9CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4ba5c413caa64fded8ff620e48a7795fed8ad4a51e14fe49c524766f7e6ccb9e","last_reissued_at":"2026-07-05T04:47:34.821981Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:47:34.821981Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2208.05321","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T04:47:34Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ykODi1H2vBdDtrloyU5zQ2YmVaDEaFnTb321pZnwNjVDNckZD2QHTfFXC3U5los6RH4Vf3LhX+Vl/Y1fki37Bg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-20T23:05:29.504489Z"},"content_sha256":"dc0ab32fc214027ff11d3c2a89146bcbd3d4da200daea16014c6a35edd7198dd","schema_version":"1.0","event_id":"sha256:dc0ab32fc214027ff11d3c2a89146bcbd3d4da200daea16014c6a35edd7198dd"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2022:JOS4IE6KUZH55WH7MIHERJ3ZL7","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"A Frequency-aware Software Cache for Large Recommendation System Embeddings","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.DC","cs.LG"],"primary_cat":"cs.IR","authors_text":"Geng Zhang, Jiarui Fang, Jiatong Han, Jin Liu, Shenggui Li, Yang You, Yongbin Li, Zhengda Bian","submitted_at":"2022-08-08T12:08:05Z","abstract_excerpt":"Deep learning recommendation models (DLRMs) have been widely applied in Internet companies. The embedding tables of DLRMs are too large to fit on GPU memory entirely. We propose a GPU-based software cache approaches to dynamically manage the embedding table in the CPU and GPU memory space by leveraging the id's frequency statistics of the target dataset. Our proposed software cache is efficient in training entire DLRMs on GPU in a synchronized update manner. It is also scaled to multiple GPUs in combination with the widely used hybrid parallel training approaches. Evaluating our prototype syst"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2208.05321","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/2208.05321/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T04:47:34Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"3loBinicONlIteY21CfA/Mj0P0uAeZuAON48vMbQ0EZ+DlqWFIrpiYOo1Upwl2Hmvy0/diQY1wI5G7p1D3R2DA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-20T23:05:29.505001Z"},"content_sha256":"08d36c3d4a5f2ffe66a29fc82b2a5534d0c6952af628e43262357d7919834faf","schema_version":"1.0","event_id":"sha256:08d36c3d4a5f2ffe66a29fc82b2a5534d0c6952af628e43262357d7919834faf"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/JOS4IE6KUZH55WH7MIHERJ3ZL7/bundle.json","state_url":"https://pith.science/pith/JOS4IE6KUZH55WH7MIHERJ3ZL7/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/JOS4IE6KUZH55WH7MIHERJ3ZL7/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-20T23:05:29Z","links":{"resolver":"https://pith.science/pith/JOS4IE6KUZH55WH7MIHERJ3ZL7","bundle":"https://pith.science/pith/JOS4IE6KUZH55WH7MIHERJ3ZL7/bundle.json","state":"https://pith.science/pith/JOS4IE6KUZH55WH7MIHERJ3ZL7/state.json","well_known_bundle":"https://pith.science/.well-known/pith/JOS4IE6KUZH55WH7MIHERJ3ZL7/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:JOS4IE6KUZH55WH7MIHERJ3ZL7","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":"92ac121773f811296471d63a2196530d6fc4fe1daaffce11d0cb15910c476efb","cross_cats_sorted":["cs.AI","cs.DC","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2022-08-08T12:08:05Z","title_canon_sha256":"ab94b7c4dd319074cb56a1f23304279c85824fcae557c2ff7d6f9eb70d27d448"},"schema_version":"1.0","source":{"id":"2208.05321","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2208.05321","created_at":"2026-07-05T04:47:34Z"},{"alias_kind":"arxiv_version","alias_value":"2208.05321v1","created_at":"2026-07-05T04:47:34Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2208.05321","created_at":"2026-07-05T04:47:34Z"},{"alias_kind":"pith_short_12","alias_value":"JOS4IE6KUZH5","created_at":"2026-07-05T04:47:34Z"},{"alias_kind":"pith_short_16","alias_value":"JOS4IE6KUZH55WH7","created_at":"2026-07-05T04:47:34Z"},{"alias_kind":"pith_short_8","alias_value":"JOS4IE6K","created_at":"2026-07-05T04:47:34Z"}],"graph_snapshots":[{"event_id":"sha256:08d36c3d4a5f2ffe66a29fc82b2a5534d0c6952af628e43262357d7919834faf","target":"graph","created_at":"2026-07-05T04:47:34Z","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/2208.05321/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Deep learning recommendation models (DLRMs) have been widely applied in Internet companies. The embedding tables of DLRMs are too large to fit on GPU memory entirely. We propose a GPU-based software cache approaches to dynamically manage the embedding table in the CPU and GPU memory space by leveraging the id's frequency statistics of the target dataset. Our proposed software cache is efficient in training entire DLRMs on GPU in a synchronized update manner. It is also scaled to multiple GPUs in combination with the widely used hybrid parallel training approaches. Evaluating our prototype syst","authors_text":"Geng Zhang, Jiarui Fang, Jiatong Han, Jin Liu, Shenggui Li, Yang You, Yongbin Li, Zhengda Bian","cross_cats":["cs.AI","cs.DC","cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2022-08-08T12:08:05Z","title":"A Frequency-aware Software Cache for Large Recommendation System Embeddings"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2208.05321","kind":"arxiv","version":1},"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:dc0ab32fc214027ff11d3c2a89146bcbd3d4da200daea16014c6a35edd7198dd","target":"record","created_at":"2026-07-05T04:47:34Z","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":"92ac121773f811296471d63a2196530d6fc4fe1daaffce11d0cb15910c476efb","cross_cats_sorted":["cs.AI","cs.DC","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2022-08-08T12:08:05Z","title_canon_sha256":"ab94b7c4dd319074cb56a1f23304279c85824fcae557c2ff7d6f9eb70d27d448"},"schema_version":"1.0","source":{"id":"2208.05321","kind":"arxiv","version":1}},"canonical_sha256":"4ba5c413caa64fded8ff620e48a7795fed8ad4a51e14fe49c524766f7e6ccb9e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4ba5c413caa64fded8ff620e48a7795fed8ad4a51e14fe49c524766f7e6ccb9e","first_computed_at":"2026-07-05T04:47:34.821981Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T04:47:34.821981Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"hQGv2MLIIdHLQed5Np43jlyIsn46NdY9kicNxvpWRHB/5lASugYsbVrkRInN6ME/0Fg+z7uAL16yKEOb43W9CQ==","signature_status":"signed_v1","signed_at":"2026-07-05T04:47:34.822451Z","signed_message":"canonical_sha256_bytes"},"source_id":"2208.05321","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:dc0ab32fc214027ff11d3c2a89146bcbd3d4da200daea16014c6a35edd7198dd","sha256:08d36c3d4a5f2ffe66a29fc82b2a5534d0c6952af628e43262357d7919834faf"],"state_sha256":"1c4c6b4f46f224513880e9f04499bb05a32b31547a7b7bbb84c4ba1b09c7c55e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"6u2fPUq7ZNVjazAg9Ep5lBebMBQvkd8Kbp7RG4lyztg62nZtx8/3Zju08Z/Hyj99SGDgTDHRyp4p/quRcu14DQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-20T23:05:29.509690Z","bundle_sha256":"e6c5e8d40085a90670fe52ca9824a5adb8a7ae267cf13ad746de561f4e371d73"}}