Flare speeds up micromagnetic simulations by predicting magnetization directly
FLARE: Flow Matching with Local Axis-Angle Representations for Stochastic Micromagnetic Evolution
Machine LearningArtificial Intelligence
Summary
Simulating how tiny magnets change over time usually takes many small calculation steps, which is slow. The researchers developed FLARE, a new way to predict how magnetization evolves directly at specific future times, skipping slow step-by-step calculations. This method respects the nature of magnetization changes and keeps the math true to physics. FLARE is shown to be much faster and more accurate than existing simulation tools on standard tests.
What this means in practice
- •For magnetic device engineers: Run faster stochastic micromagnetic simulations for design and testing at multiple future times without stepwise calculations.
- •For gpu software developers: Integrate FLARE's flow-matching approach to accelerate existing GPU micromagnetic solvers like MuMax3 significantly.
Authors
Pengyu Li, Renjie Tong, Xuanlue Jiang, Jianmin Li, Yuanyuan Zhou
Abstract
Long-horizon micromagnetic simulation remains expensive because conventional and learned solvers typically propagate Landau--Lifshitz--Gilbert (LLG) dynamics step by step. Existing learned approaches generally retain stepwise integration or model deterministic evolution, leaving full-field, direct-horizon stochastic prediction largely unexplored. We propose FLARE, a flow-matching framework that recasts stochastic finite-time magnetization prediction as conditional transport over anchor-relative local axis-angle rotations. This rotation-space formulation respects the intrinsic geometry of magnetization dynamics and preserves pointwise unit norm by construction. By explicitly conditioning on the physical prediction horizon, FLARE directly generates full-field stochastic endpoints across multiple target times without stepwise integration. Against the strongest single-checkpoint external baseline on each metric, FLARE achieves 29.9% lower angular energy distance ($15.30^\circ$), and a 37.3% lower fair energy score (0.393). On a representative composed 5-ns two-segment protocol, FLARE achieves a $3{,}062\times$ best-batch speedup over the widely used GPU micromagnetic solver MuMax$^3$ on a single GPU.