{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:UIHZW6WONVRS7JSCH6EIS72FLN","short_pith_number":"pith:UIHZW6WO","schema_version":"1.0","canonical_sha256":"a20f9b7ace6d632fa6423f88897f455b5b53c0e60e4bf91b492987681adc1eb7","source":{"kind":"arxiv","id":"2501.10546","version":1},"attestation_state":"computed","paper":{"title":"Scalable Machine Learning Training Infrastructure for Online Ads Recommendation and Auction Scoring Modeling at Google","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.DC","authors_text":"Abdulrahman Salem, Felix Berger, Gary Holt, George Kurian, Herve Quiroz, Jeremiah Willcock, Julian Grady, Kaiyuan Wang, Ryan Sims, Somayeh Sardashti, Yang Li","submitted_at":"2025-01-17T20:40:56Z","abstract_excerpt":"Large-scale Ads recommendation and auction scoring models at Google scale demand immense computational resources. While specialized hardware like TPUs have improved linear algebra computations, bottlenecks persist in large-scale systems. This paper proposes solutions for three critical challenges that must be addressed for efficient end-to-end execution in a widely used production infrastructure: (1) Input Generation and Ingestion Pipeline: Efficiently transforming raw features (e.g., \"search query\") into numerical inputs and streaming them to TPUs; (2) Large Embedding Tables: Optimizing conve"},"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":"2501.10546","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2025-01-17T20:40:56Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"e1d01e6b2d85c6fcb2f2f543ff93380df83737f4813d32ecf6c811116948d609","abstract_canon_sha256":"6126df6d5ce9842c99103d91de5813b1ceae46832cf028fc6af6388352b24c7f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:02:29.650945Z","signature_b64":"7SgBwJN0rZi46YrdYK65Q61VeC+CnwcY0KYKAKzG2cxMboQnJxz8fKcJauCtdhcEKGD8/rS3EFtinI5x8F/ZBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a20f9b7ace6d632fa6423f88897f455b5b53c0e60e4bf91b492987681adc1eb7","last_reissued_at":"2026-07-05T10:02:29.650442Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:02:29.650442Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Scalable Machine Learning Training Infrastructure for Online Ads Recommendation and Auction Scoring Modeling at Google","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.DC","authors_text":"Abdulrahman Salem, Felix Berger, Gary Holt, George Kurian, Herve Quiroz, Jeremiah Willcock, Julian Grady, Kaiyuan Wang, Ryan Sims, Somayeh Sardashti, Yang Li","submitted_at":"2025-01-17T20:40:56Z","abstract_excerpt":"Large-scale Ads recommendation and auction scoring models at Google scale demand immense computational resources. While specialized hardware like TPUs have improved linear algebra computations, bottlenecks persist in large-scale systems. This paper proposes solutions for three critical challenges that must be addressed for efficient end-to-end execution in a widely used production infrastructure: (1) Input Generation and Ingestion Pipeline: Efficiently transforming raw features (e.g., \"search query\") into numerical inputs and streaming them to TPUs; (2) Large Embedding Tables: Optimizing conve"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.10546","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/2501.10546/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":"2501.10546","created_at":"2026-07-05T10:02:29.650501+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.10546v1","created_at":"2026-07-05T10:02:29.650501+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.10546","created_at":"2026-07-05T10:02:29.650501+00:00"},{"alias_kind":"pith_short_12","alias_value":"UIHZW6WONVRS","created_at":"2026-07-05T10:02:29.650501+00:00"},{"alias_kind":"pith_short_16","alias_value":"UIHZW6WONVRS7JSC","created_at":"2026-07-05T10:02:29.650501+00:00"},{"alias_kind":"pith_short_8","alias_value":"UIHZW6WO","created_at":"2026-07-05T10:02:29.650501+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/UIHZW6WONVRS7JSCH6EIS72FLN","json":"https://pith.science/pith/UIHZW6WONVRS7JSCH6EIS72FLN.json","graph_json":"https://pith.science/api/pith-number/UIHZW6WONVRS7JSCH6EIS72FLN/graph.json","events_json":"https://pith.science/api/pith-number/UIHZW6WONVRS7JSCH6EIS72FLN/events.json","paper":"https://pith.science/paper/UIHZW6WO"},"agent_actions":{"view_html":"https://pith.science/pith/UIHZW6WONVRS7JSCH6EIS72FLN","download_json":"https://pith.science/pith/UIHZW6WONVRS7JSCH6EIS72FLN.json","view_paper":"https://pith.science/paper/UIHZW6WO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.10546&json=true","fetch_graph":"https://pith.science/api/pith-number/UIHZW6WONVRS7JSCH6EIS72FLN/graph.json","fetch_events":"https://pith.science/api/pith-number/UIHZW6WONVRS7JSCH6EIS72FLN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UIHZW6WONVRS7JSCH6EIS72FLN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UIHZW6WONVRS7JSCH6EIS72FLN/action/storage_attestation","attest_author":"https://pith.science/pith/UIHZW6WONVRS7JSCH6EIS72FLN/action/author_attestation","sign_citation":"https://pith.science/pith/UIHZW6WONVRS7JSCH6EIS72FLN/action/citation_signature","submit_replication":"https://pith.science/pith/UIHZW6WONVRS7JSCH6EIS72FLN/action/replication_record"}},"created_at":"2026-07-05T10:02:29.650501+00:00","updated_at":"2026-07-05T10:02:29.650501+00:00"}