A parallel-in-time approach for accelerating direct-adjoint studies
arXiv:2004.00546 · doi:10.1016/j.jcp.2020.110033
Abstract
Parallel-in-time methods are developed to accelerate the direct-adjoint looping procedure. Particularly, we utilize the Paraexp algorithm, previously developed to integrate equations forward in time, to accelerate the direct-adjoint looping that arises from gradient-based optimization. We consider both linear and non-linear governing equations and exploit the linear, time-varying nature of the adjoint equations. Gains in efficiency are seen across all cases, showing that a Paraexp based parallel-in-time approach is feasible for the acceleration of direct-adjoint studies. This signifies a possible approach to further increase the run-time performance for optimization studies that either cannot be parallelized in space or are at their limit of efficiency gains for a parallel-in-space approach.
arXiv comment37 pages, 9 figures
References in corpus (10)
- P3DFFT: a framework for parallel computations of Fourier transforms in three dimensions
- Mixing enhancement in binary fluids using optimised stirring strategies
- An Efficient Parallel-in-Time Method for Optimization with Parabolic PDEs
- ParaDiag: parallel-in-time algorithms based on the diagonalization technique
- Optimal mixing in two-dimensional stratified plane Poiseuille flow at finite Péclet and Richardson numbers
- Limited-memory polynomial methods for large-scale matrix functions
- A generalized massively parallel ultra-high order FFT-based Maxwell solver
- A gradient-based framework for maximizing mixing in binary fluids
- Performance of parallel-in-time integration for Rayleigh Bénard Convection
- Shape optimisation of stirring rods in mixing binary fluids