Neural operators beat classical solvers at low accuracy but not high
Cost-Accuracy Trade-offs: Neural Operator vs Classical Numerical Solver
Machine Learning
Summary
Solving complex physical problems often requires repeated calculations for different inputs. The authors study when fast, AI-based neural operator models can replace slower traditional computer methods called classical numerical solvers. They find neural operators work well and save time when a rough but quick answer is good enough. But for very precise answers, the traditional methods still perform better. This helps decide which method to use depending on how accurate and fast the solution must be.
What this means in practice
- •For engineering simulation teams: Speed up repeated engineering simulations by using neural operators when moderate accuracy suffices, saving computational time and cost.
- •For financial modelers: Use neural operators to quickly estimate outcomes in models involving differential equations when exact precision is less critical.
Authors
Daniel Zhengyu Huang, Andrew M. Stuart
Abstract
Neural operators are data-driven models that learn mappings from inputs that parameterize partial differential equations, such as spatially varying coefficients, initial conditions, forcing terms, boundary conditions, or geometries, to solution fields or quantities of interest. Once trained, they can serve as surrogates for classical numerical solvers in many-query settings that require repeated evaluations for varying inputs. We address the question of when, and then why, neural operator surrogates outperform classical numerical solvers, in terms of cost for a given accuracy. We focus on the post-training, many-query limit, in which data-acquisition and training costs are treated as fixed and fully amortized. Even in this deliberately favorable regime for neural operators, there are regimes in which classical solvers outperform the surrogate models. We compare the cost-accuracy performance of neural operator surrogates and classical numerical solvers through a reproducible benchmark study comparing neural operators with problem-matched classical solvers on representative problems in computational science and engineering, focusing on prediction error, per-query floating-point cost, and wall-clock runtime. Neural operators are most competitive at low-to-moderate accuracy requirements. Their floating-point cost advantage depends strongly on the problem structure, arising when they avoid temporal or nonlinear iterations or predict a reduced quantity of interest rather than a full solution field. Additional wall-clock speedups result from dense tensor operations that are well suited to modern hardware. As the target accuracy is tightened, achieving the required accuracy with neural operators becomes increasingly challenging, and classical solvers outperform surrogates in this regime; thus classical solvers will remain important for verification and high-accuracy computation.