{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2019:BMNLQDULGWOOHZNQIPAEU7TJXP","short_pith_number":"pith:BMNLQDUL","canonical_record":{"source":{"id":"1904.06376","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.MS","submitted_at":"2019-04-12T19:04:47Z","cross_cats_sorted":["cs.NA"],"title_canon_sha256":"a2d05c8d21c614040ae0112606f062bf801bced6e4380e5c6df2af9a09afc3cd","abstract_canon_sha256":"d16b3cc640ca196581c299c112db765c81f0768e80eddc4ae08ec0b0e9f71158"},"schema_version":"1.0"},"canonical_sha256":"0b1ab80e8b359ce3e5b043c04a7e69bbfabb0cdaa3cce1755648321e7711e766","source":{"kind":"arxiv","id":"1904.06376","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1904.06376","created_at":"2026-05-17T23:48:41Z"},{"alias_kind":"arxiv_version","alias_value":"1904.06376v1","created_at":"2026-05-17T23:48:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1904.06376","created_at":"2026-05-17T23:48:41Z"},{"alias_kind":"pith_short_12","alias_value":"BMNLQDULGWOO","created_at":"2026-05-18T12:33:12Z"},{"alias_kind":"pith_short_16","alias_value":"BMNLQDULGWOOHZNQ","created_at":"2026-05-18T12:33:12Z"},{"alias_kind":"pith_short_8","alias_value":"BMNLQDUL","created_at":"2026-05-18T12:33:12Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2019:BMNLQDULGWOOHZNQIPAEU7TJXP","target":"record","payload":{"canonical_record":{"source":{"id":"1904.06376","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.MS","submitted_at":"2019-04-12T19:04:47Z","cross_cats_sorted":["cs.NA"],"title_canon_sha256":"a2d05c8d21c614040ae0112606f062bf801bced6e4380e5c6df2af9a09afc3cd","abstract_canon_sha256":"d16b3cc640ca196581c299c112db765c81f0768e80eddc4ae08ec0b0e9f71158"},"schema_version":"1.0"},"canonical_sha256":"0b1ab80e8b359ce3e5b043c04a7e69bbfabb0cdaa3cce1755648321e7711e766","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:48:41.495488Z","signature_b64":"FqLyDob7XHeAfv/ycHlDo19OJyEb2ZdzbkB7SE3+75OkFqU251p25JUCSdH8GBw+hWCRlDcmwrVLvJKAWj3uBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0b1ab80e8b359ce3e5b043c04a7e69bbfabb0cdaa3cce1755648321e7711e766","last_reissued_at":"2026-05-17T23:48:41.495004Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:48:41.495004Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1904.06376","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":"2026-05-17T23:48:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IuaJT5u7RNhXoc7P4Q4kzziriMDiNQ6JbLKJlaH5f/vlEWEzEw5N3wubnDwGVTGha1FLsFtQzLeRTNizYc4YAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-28T19:22:00.606013Z"},"content_sha256":"8d67dfb0bd118f9ccd389f08b9b6de6543960b2a6f9f3a352b64d2f9f633fb11","schema_version":"1.0","event_id":"sha256:8d67dfb0bd118f9ccd389f08b9b6de6543960b2a6f9f3a352b64d2f9f633fb11"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2019:BMNLQDULGWOOHZNQIPAEU7TJXP","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Leveraging the bfloat16 Artificial Intelligence Datatype For Higher-Precision Computations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NA"],"primary_cat":"cs.MS","authors_text":"Alexander Heinecke, Greg Henry, Ping Tak Peter Tang","submitted_at":"2019-04-12T19:04:47Z","abstract_excerpt":"In recent years fused-multiply-add (FMA) units with lower-precision multiplications and higher-precision accumulation have proven useful in machine learning/artificial intelligence applications, most notably in training deep neural networks due to their extreme computational intensity. Compared to classical IEEE-754 32 bit (FP32) and 64 bit (FP64) arithmetic, these reduced precision arithmetic can naturally be sped up disproportional to their shortened width. The common strategy of all major hardware vendors is to aggressively further enhance their performance disproportionately. One particula"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1904.06376","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":""},"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":"2026-05-17T23:48:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"c6UZsfAKw5IkwDjigKRfeziKG3rmAnYgNCktljDUMJHTSwI6vJe16sXSnGiZB0NbpS1X92elZe2cZAmKz1KADA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-28T19:22:00.606728Z"},"content_sha256":"81c261ecf4a351e5f4d7d0fd0fa5d6418bb78d00fda3657dab16cfb73f239b59","schema_version":"1.0","event_id":"sha256:81c261ecf4a351e5f4d7d0fd0fa5d6418bb78d00fda3657dab16cfb73f239b59"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/BMNLQDULGWOOHZNQIPAEU7TJXP/bundle.json","state_url":"https://pith.science/pith/BMNLQDULGWOOHZNQIPAEU7TJXP/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/BMNLQDULGWOOHZNQIPAEU7TJXP/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":"2026-05-28T19:22:00Z","links":{"resolver":"https://pith.science/pith/BMNLQDULGWOOHZNQIPAEU7TJXP","bundle":"https://pith.science/pith/BMNLQDULGWOOHZNQIPAEU7TJXP/bundle.json","state":"https://pith.science/pith/BMNLQDULGWOOHZNQIPAEU7TJXP/state.json","well_known_bundle":"https://pith.science/.well-known/pith/BMNLQDULGWOOHZNQIPAEU7TJXP/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2019:BMNLQDULGWOOHZNQIPAEU7TJXP","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":"d16b3cc640ca196581c299c112db765c81f0768e80eddc4ae08ec0b0e9f71158","cross_cats_sorted":["cs.NA"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.MS","submitted_at":"2019-04-12T19:04:47Z","title_canon_sha256":"a2d05c8d21c614040ae0112606f062bf801bced6e4380e5c6df2af9a09afc3cd"},"schema_version":"1.0","source":{"id":"1904.06376","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1904.06376","created_at":"2026-05-17T23:48:41Z"},{"alias_kind":"arxiv_version","alias_value":"1904.06376v1","created_at":"2026-05-17T23:48:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1904.06376","created_at":"2026-05-17T23:48:41Z"},{"alias_kind":"pith_short_12","alias_value":"BMNLQDULGWOO","created_at":"2026-05-18T12:33:12Z"},{"alias_kind":"pith_short_16","alias_value":"BMNLQDULGWOOHZNQ","created_at":"2026-05-18T12:33:12Z"},{"alias_kind":"pith_short_8","alias_value":"BMNLQDUL","created_at":"2026-05-18T12:33:12Z"}],"graph_snapshots":[{"event_id":"sha256:81c261ecf4a351e5f4d7d0fd0fa5d6418bb78d00fda3657dab16cfb73f239b59","target":"graph","created_at":"2026-05-17T23:48:41Z","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"},"paper":{"abstract_excerpt":"In recent years fused-multiply-add (FMA) units with lower-precision multiplications and higher-precision accumulation have proven useful in machine learning/artificial intelligence applications, most notably in training deep neural networks due to their extreme computational intensity. Compared to classical IEEE-754 32 bit (FP32) and 64 bit (FP64) arithmetic, these reduced precision arithmetic can naturally be sped up disproportional to their shortened width. The common strategy of all major hardware vendors is to aggressively further enhance their performance disproportionately. One particula","authors_text":"Alexander Heinecke, Greg Henry, Ping Tak Peter Tang","cross_cats":["cs.NA"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.MS","submitted_at":"2019-04-12T19:04:47Z","title":"Leveraging the bfloat16 Artificial Intelligence Datatype For Higher-Precision Computations"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1904.06376","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":"sha256:8d67dfb0bd118f9ccd389f08b9b6de6543960b2a6f9f3a352b64d2f9f633fb11","target":"record","created_at":"2026-05-17T23:48:41Z","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":"d16b3cc640ca196581c299c112db765c81f0768e80eddc4ae08ec0b0e9f71158","cross_cats_sorted":["cs.NA"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.MS","submitted_at":"2019-04-12T19:04:47Z","title_canon_sha256":"a2d05c8d21c614040ae0112606f062bf801bced6e4380e5c6df2af9a09afc3cd"},"schema_version":"1.0","source":{"id":"1904.06376","kind":"arxiv","version":1}},"canonical_sha256":"0b1ab80e8b359ce3e5b043c04a7e69bbfabb0cdaa3cce1755648321e7711e766","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"0b1ab80e8b359ce3e5b043c04a7e69bbfabb0cdaa3cce1755648321e7711e766","first_computed_at":"2026-05-17T23:48:41.495004Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:48:41.495004Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"FqLyDob7XHeAfv/ycHlDo19OJyEb2ZdzbkB7SE3+75OkFqU251p25JUCSdH8GBw+hWCRlDcmwrVLvJKAWj3uBw==","signature_status":"signed_v1","signed_at":"2026-05-17T23:48:41.495488Z","signed_message":"canonical_sha256_bytes"},"source_id":"1904.06376","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:8d67dfb0bd118f9ccd389f08b9b6de6543960b2a6f9f3a352b64d2f9f633fb11","sha256:81c261ecf4a351e5f4d7d0fd0fa5d6418bb78d00fda3657dab16cfb73f239b59"],"state_sha256":"2d46b8856d7085e1eac232dc9ceb78a98cfa64f7fdc4a8c00db4863f3dd97736"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"+FlevXfXJPy0qQbf3eLVLm8wEY/X0I2/8gFVdyywAVamU6b/rDWUZaFqPUQAfGklnkpWURoeHSc7sZmEUsjXAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-28T19:22:00.610435Z","bundle_sha256":"c644c9cd4deb6a59f6277436261b5382e36692c75c5a105025bdcaa49f5a9ca6"}}