{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:TBLPLPCNW2TZMMPFUQZURZTF4D","short_pith_number":"pith:TBLPLPCN","schema_version":"1.0","canonical_sha256":"9856f5bc4db6a79631e5a43348e665e0f47940f346e53e8de30418d42d16c547","source":{"kind":"arxiv","id":"2507.14715","version":1},"attestation_state":"computed","paper":{"title":"Exploring the Dynamic Scheduling Space of Real-Time Generative AI Applications on Emerging Heterogeneous Systems","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Ashish Sirasao, Hyoukjun Kwon, Rachid Karami, Rajeev Patwari","submitted_at":"2025-07-19T18:24:11Z","abstract_excerpt":"The integration of generative AI models, particularly large language models (LLMs), into real-time multi-model AI applications such as video conferencing and gaming is giving rise to a new class of workloads: real-time generative AI (RTGen). These workloads combine the compute intensity and dynamic execution patterns of generative models with the stringent latency and concurrency constraints of real-time inference. To meet the diverse demands of RTGen workloads, modern edge platforms increasingly adopt heterogeneous system-on-chip (SoC) architectures that integrate CPUs, GPUs, and NPUs. Despit"},"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":"2507.14715","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-07-19T18:24:11Z","cross_cats_sorted":[],"title_canon_sha256":"92893a820de8ac1fadb1c117d856a7bbb105d0bb3f9c532430d58b90a66454fa","abstract_canon_sha256":"3bc6544cb130be6d041062efd08c133d24416d3a7472ac65848f8e6c20793ce8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:40:12.349693Z","signature_b64":"Bstib8aVTRqlz4fJdMRkKHi3hz5GYyiCfUxhzRpnV/UdnqVnlBoJcPjXC4wzpb/UkZqf1msX3GRbZCnyTm8hAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9856f5bc4db6a79631e5a43348e665e0f47940f346e53e8de30418d42d16c547","last_reissued_at":"2026-07-05T11:40:12.349119Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:40:12.349119Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Exploring the Dynamic Scheduling Space of Real-Time Generative AI Applications on Emerging Heterogeneous Systems","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Ashish Sirasao, Hyoukjun Kwon, Rachid Karami, Rajeev Patwari","submitted_at":"2025-07-19T18:24:11Z","abstract_excerpt":"The integration of generative AI models, particularly large language models (LLMs), into real-time multi-model AI applications such as video conferencing and gaming is giving rise to a new class of workloads: real-time generative AI (RTGen). These workloads combine the compute intensity and dynamic execution patterns of generative models with the stringent latency and concurrency constraints of real-time inference. To meet the diverse demands of RTGen workloads, modern edge platforms increasingly adopt heterogeneous system-on-chip (SoC) architectures that integrate CPUs, GPUs, and NPUs. Despit"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.14715","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/2507.14715/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":"2507.14715","created_at":"2026-07-05T11:40:12.349183+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.14715v1","created_at":"2026-07-05T11:40:12.349183+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.14715","created_at":"2026-07-05T11:40:12.349183+00:00"},{"alias_kind":"pith_short_12","alias_value":"TBLPLPCNW2TZ","created_at":"2026-07-05T11:40:12.349183+00:00"},{"alias_kind":"pith_short_16","alias_value":"TBLPLPCNW2TZMMPF","created_at":"2026-07-05T11:40:12.349183+00:00"},{"alias_kind":"pith_short_8","alias_value":"TBLPLPCN","created_at":"2026-07-05T11:40:12.349183+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.29262","citing_title":"Harmonizing Real-Time Constraints and Long-Horizon Reasoning: An Asynchronous Agentic Framework for Dynamic Scheduling","ref_index":13,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TBLPLPCNW2TZMMPFUQZURZTF4D","json":"https://pith.science/pith/TBLPLPCNW2TZMMPFUQZURZTF4D.json","graph_json":"https://pith.science/api/pith-number/TBLPLPCNW2TZMMPFUQZURZTF4D/graph.json","events_json":"https://pith.science/api/pith-number/TBLPLPCNW2TZMMPFUQZURZTF4D/events.json","paper":"https://pith.science/paper/TBLPLPCN"},"agent_actions":{"view_html":"https://pith.science/pith/TBLPLPCNW2TZMMPFUQZURZTF4D","download_json":"https://pith.science/pith/TBLPLPCNW2TZMMPFUQZURZTF4D.json","view_paper":"https://pith.science/paper/TBLPLPCN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.14715&json=true","fetch_graph":"https://pith.science/api/pith-number/TBLPLPCNW2TZMMPFUQZURZTF4D/graph.json","fetch_events":"https://pith.science/api/pith-number/TBLPLPCNW2TZMMPFUQZURZTF4D/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TBLPLPCNW2TZMMPFUQZURZTF4D/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TBLPLPCNW2TZMMPFUQZURZTF4D/action/storage_attestation","attest_author":"https://pith.science/pith/TBLPLPCNW2TZMMPFUQZURZTF4D/action/author_attestation","sign_citation":"https://pith.science/pith/TBLPLPCNW2TZMMPFUQZURZTF4D/action/citation_signature","submit_replication":"https://pith.science/pith/TBLPLPCNW2TZMMPFUQZURZTF4D/action/replication_record"}},"created_at":"2026-07-05T11:40:12.349183+00:00","updated_at":"2026-07-05T11:40:12.349183+00:00"}