{"id":"W4415707388","doi":"10.1109/jproc.2025.3623023","title":"Field-Programmable Gate Array Architecture for Deep Learning: Survey and Future Directions","year":2025,"lang":"","type":"article","venue":"Proceedings of the IEEE","topic":"Advanced Neural Network Applications","field":"Computer Science","cited_by":10,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto; University of Waterloo","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Field-programmable gate array; Interfacing; Dataflow; Variety (cybernetics); Architecture; Dataflow architecture; Reconfigurable computing; Data processing; Graphics; Emulation","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0004752251,0.0002962236,0.0003464599,0.0001219158,0.0009789453,0.0002412923,0.001210935,0.0001787248,0.00000288579],"category_scores_gemma":[0.0003484691,0.0002370413,0.000172497,0.001714689,0.0001851067,0.0002711808,0.0002968045,0.00064528,0.000001062264],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003698121,"about_ca_system_score_gemma":0.00005408809,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002589032,"about_ca_topic_score_gemma":0.00007410567,"domain_scores_codex":[0.9981573,0.0000284777,0.0004080372,0.0006991885,0.000210548,0.0004964813],"domain_scores_gemma":[0.9980178,0.0006112264,0.000404474,0.0002978625,0.0005696834,0.00009888948],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0003357016,0.0002271481,0.01032979,0.0008236959,0.0002398887,1.339715e-7,0.002694592,0.001785434,0.01832485,0.02555123,0.005167839,0.9345197],"study_design_scores_gemma":[0.001370568,0.0007524145,0.01353448,0.0004532062,0.0002441145,0.00003241063,0.0003839702,0.05411512,0.09663653,0.09480449,0.7367141,0.0009585829],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1633497,0.01043218,0.7402304,0.06602707,0.006380053,0.006554684,0.00004134189,0.0005708846,0.006413678],"genre_scores_gemma":[0.9231207,0.002956679,0.0642724,0.0008994422,0.001041991,0.0006415774,0.000002945471,0.00005898566,0.007005225],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9335611,"threshold_uncertainty_score":0.9666263,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01311120761162512,"score_gpt":0.2634823301032785,"score_spread":0.2503711224916534,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}