{"paper":{"title":"JAX-Based Batched AC Power Flow for GPU Acceleration and AI Ecosystem Integration","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"A batched implementation of AC power flow solvers on accelerators achieves over 10 times the speed of existing methods and integrates directly with AI techniques for power system management.","cross_cats":["cs.SY"],"primary_cat":"eess.SY","authors_text":"Dylan Cope, Jakob Foerster, Thomas Morstyn, Yihong Zhou","submitted_at":"2026-05-13T20:42:06Z","abstract_excerpt":"Coordinating growing grid flexibility under uncertainty is becoming increasingly important for efficient and reliable power-system operation. A core computational requirement is the efficient large-scale batched evaluation of AC power flow across candidate operating actions and uncertainty scenarios. Previous work has explored GPU-based batched power-flow evaluation, but has largely relied on hand-written C or CUDA code, creating barriers to customisation, efficient kernel optimisation, and long-term maintenance. JAX is a Python-based framework that enables efficient accelerator execution whil"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"achieving more than 10x speed-ups relative to pandapower and OpenDSS... making it straightforward to embed power-flow evaluation within AI methods for future larger-scale and more complex power-system operation","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the JAX implementations of Newton-Raphson and Z-Bus exactly reproduce the numerical behaviour and convergence properties of the reference solvers across realistic network sizes and operating conditions without hidden accuracy loss or instability.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"A JAX-based batched AC power flow solver delivers over 10x speedups versus pandapower and OpenDSS while integrating directly with AI ecosystems for power-system operation.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A batched implementation of AC power flow solvers on accelerators achieves over 10 times the speed of existing methods and integrates directly with AI techniques for power system management.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"101bbfdac5e768d779a5ed26a0c1b6a6e5b73f9172f2e1cf905fd1098e098290"},"source":{"id":"2605.14103","kind":"arxiv","version":1},"verdict":{"id":"c23b2902-30b3-433f-a843-589b3845d712","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T05:10:47.899787Z","strongest_claim":"achieving more than 10x speed-ups relative to pandapower and OpenDSS... making it straightforward to embed power-flow evaluation within AI methods for future larger-scale and more complex power-system operation","one_line_summary":"A JAX-based batched AC power flow solver delivers over 10x speedups versus pandapower and OpenDSS while integrating directly with AI ecosystems for power-system operation.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the JAX implementations of Newton-Raphson and Z-Bus exactly reproduce the numerical behaviour and convergence properties of the reference solvers across realistic network sizes and operating conditions without hidden accuracy loss or instability.","pith_extraction_headline":"A batched implementation of AC power flow solvers on accelerators achieves over 10 times the speed of existing methods and integrates directly with AI techniques for power system management."},"references":{"count":12,"sample":[{"doi":"","year":2018,"title":"2018 , url =","work_id":"1b73edb8-c3ef-432c-8491-10214fc523a0","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Pandapower—An Open-Source Python Tool for Convenient Modeling, Analysis, and Optimization of Electric Power Systems , year=","work_id":"88f76d27-8ad1-4951-b30f-f90e9d8a10f0","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2018,"title":"2018 , url =","work_id":"e1b5e88a-7ac6-4bc8-b7a2-a96b52b7147f","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Network Topology Optimisation , year =","work_id":"d3902950-fdc6-470c-8b1a-965533f216e2","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Stott, B. and Alsac, O. , journal=. Fast Decoupled Load Flow , year=","work_id":"201477fe-f318-4b96-9399-f2fda5db9304","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":12,"snapshot_sha256":"70de3e0738c307ed6c261bcb46e94c4d31f4d5a45dec4230396d5c5d57bec45b","internal_anchors":0},"formal_canon":{"evidence_count":1,"snapshot_sha256":"47ee6d91aa837a85ec9f540a3279aef97d14d1c002d1dfe08bd3e1bd473dfda7"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}