安裝中文字典英文字典辭典工具!
安裝中文字典英文字典辭典工具!
|
- Data-driven acceleration of photonic simulations | Scientific . . .
Here we develop an accelerated solver (data-driven GMRES) by interfacing a machine learning model with GMRES The machine learning model is trained to predict a subspace that approximates the
- arXiv:2110. 12444v1 [physics. plasm-ph] 24 Oct 2021
this contribution we explore how machine learning can be used to minimize the number of GMRES iterations required to advance the PIC system in time In particular, Neural Networks are trained to suggest vectors which augment the initial Krylov space of GMRES iterations [13, 14, 15] Interfacing
- Accelerating GMRES with Deep Learning in Real-Time - SC20
GMRES is a popular solver for the Laplace equation (a common PDE in physical systems) We demonstrate that our method can efficiently train convolutional neural networks to produce tuned preconditioners for GMRES solvers, that provide a roughly 2 acceleration
- Graph neural network-based preconditioners for optimizing . . .
Finding suitable preconditioners to accelerate iterative solution methods remains an active area of research In this project, we address the challenge of accelerating GMRES, a Krylov subspace method for solving large-scale linear systems, by developing a data-driven preconditioner
- arXiv:1902. 00090v3 [physics. comp-ph] 25 Mar 2020
Here we develop an accelerated solver (data-driven GMRES) by interfacing a machine learning model with GMRES The machine learning model is trained to predict a subspace that approximates the simulation result, and the subspace is used to augment the GMRES iterations
- Graph Neural Networks (GNNs) based accelerated numerical . . .
In this work, a novel surrogate model based on graph neural networks (GNNs) is proposed and it can accurately predict stress distribution for any given mesh structure or geometry
- An efficient GPU version of the preconditioned GMRES method
In this section, we describe the main aspects of the techniques employed to accelerate the GPU variant of the GMRES method in ILUPACK First, we detail our efforts to accelerate the MGSO procedure using the GPU, to then proceed to describe the new SpTrSV routine that we employ seeking to enhance the performance of the preconditioner
|
|
|