{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:SONK3ZI7ZNHGY6YFP4DDRTECL3","short_pith_number":"pith:SONK3ZI7","schema_version":"1.0","canonical_sha256":"939aade51fcb4e6c7b057f0638cc825ef32a8389af15ad2b4f6a4f305f545b1d","source":{"kind":"arxiv","id":"2507.20420","version":1},"attestation_state":"computed","paper":{"title":"Demystifying the 7-D Convolution Loop Nest for Data and Instruction Streaming in Reconfigurable AI Accelerators","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AR","authors_text":"Md Rownak Hossain Chowdhury, Mostafizur Rahman","submitted_at":"2025-07-27T21:33:34Z","abstract_excerpt":"Convolution remains the most compute-intensive operation in AI acceleration, often constituting over 80-90% of the workload. Existing approaches in spatial architectures such as coarse-grained reconfigurable arrays (CGRAs) and field-programmable gate arrays (FPGAs) frequently rely on loop unrolling or GEMM-based matrix transformations, introducing significant overhead in both data movement and instruction control. This paper presents a new framework designed to systematically demystify the 7-dimensional convolution loop nest by reinterpreting it as a hardware-centric data and instruction strea"},"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":"2507.20420","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AR","submitted_at":"2025-07-27T21:33:34Z","cross_cats_sorted":[],"title_canon_sha256":"4d30d925b49ac6e85aab07a490b385511d40b0d68bfd3680d20ded56b382f35a","abstract_canon_sha256":"12c17487e94ad3610ef9ff0d96dd2ae855f16705e390f1ed01b43da299f66064"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:44:19.136512Z","signature_b64":"MfxFl004boqCP6AwmWSxTj7XDpV62Hm4AiLscWAN5dXSEUEMrnLxtjnvnyrxrgzrbAL+KQz6S2fQTxzk4qZkDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"939aade51fcb4e6c7b057f0638cc825ef32a8389af15ad2b4f6a4f305f545b1d","last_reissued_at":"2026-07-05T11:44:19.136069Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:44:19.136069Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Demystifying the 7-D Convolution Loop Nest for Data and Instruction Streaming in Reconfigurable AI Accelerators","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AR","authors_text":"Md Rownak Hossain Chowdhury, Mostafizur Rahman","submitted_at":"2025-07-27T21:33:34Z","abstract_excerpt":"Convolution remains the most compute-intensive operation in AI acceleration, often constituting over 80-90% of the workload. Existing approaches in spatial architectures such as coarse-grained reconfigurable arrays (CGRAs) and field-programmable gate arrays (FPGAs) frequently rely on loop unrolling or GEMM-based matrix transformations, introducing significant overhead in both data movement and instruction control. This paper presents a new framework designed to systematically demystify the 7-dimensional convolution loop nest by reinterpreting it as a hardware-centric data and instruction strea"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.20420","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/2507.20420/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":"2507.20420","created_at":"2026-07-05T11:44:19.136127+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.20420v1","created_at":"2026-07-05T11:44:19.136127+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.20420","created_at":"2026-07-05T11:44:19.136127+00:00"},{"alias_kind":"pith_short_12","alias_value":"SONK3ZI7ZNHG","created_at":"2026-07-05T11:44:19.136127+00:00"},{"alias_kind":"pith_short_16","alias_value":"SONK3ZI7ZNHGY6YF","created_at":"2026-07-05T11:44:19.136127+00:00"},{"alias_kind":"pith_short_8","alias_value":"SONK3ZI7","created_at":"2026-07-05T11:44:19.136127+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2509.03846","citing_title":"Hardware-Aware Data and Instruction Mapping for AI Tasks: Balancing Parallelism, I/O and Memory Tradeoffs","ref_index":42,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SONK3ZI7ZNHGY6YFP4DDRTECL3","json":"https://pith.science/pith/SONK3ZI7ZNHGY6YFP4DDRTECL3.json","graph_json":"https://pith.science/api/pith-number/SONK3ZI7ZNHGY6YFP4DDRTECL3/graph.json","events_json":"https://pith.science/api/pith-number/SONK3ZI7ZNHGY6YFP4DDRTECL3/events.json","paper":"https://pith.science/paper/SONK3ZI7"},"agent_actions":{"view_html":"https://pith.science/pith/SONK3ZI7ZNHGY6YFP4DDRTECL3","download_json":"https://pith.science/pith/SONK3ZI7ZNHGY6YFP4DDRTECL3.json","view_paper":"https://pith.science/paper/SONK3ZI7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.20420&json=true","fetch_graph":"https://pith.science/api/pith-number/SONK3ZI7ZNHGY6YFP4DDRTECL3/graph.json","fetch_events":"https://pith.science/api/pith-number/SONK3ZI7ZNHGY6YFP4DDRTECL3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SONK3ZI7ZNHGY6YFP4DDRTECL3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SONK3ZI7ZNHGY6YFP4DDRTECL3/action/storage_attestation","attest_author":"https://pith.science/pith/SONK3ZI7ZNHGY6YFP4DDRTECL3/action/author_attestation","sign_citation":"https://pith.science/pith/SONK3ZI7ZNHGY6YFP4DDRTECL3/action/citation_signature","submit_replication":"https://pith.science/pith/SONK3ZI7ZNHGY6YFP4DDRTECL3/action/replication_record"}},"created_at":"2026-07-05T11:44:19.136127+00:00","updated_at":"2026-07-05T11:44:19.136127+00:00"}