Scalable Finite-Element Viscoelastic Crustal Deformation Analysis Accelerated with Data-Driven Method
Scalable Finite-Element Viscoelastic Crustal Deformation Analysis Accelerated with Data-Driven Method
复制标题
利用数据驱动方法加速可扩展有限元粘弹性地壳变形分析
DOI：
10.1109/scalah56622.2022.00008
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
Ueda Naonori
中科院分区：
文献类型：
--
作者：
Fujita Kohei;Murakami Sota;Ichimura Tsuyoshi;Hori Takane;Hori Muneo;Lalith Maddegedara;Ueda Naonori
针对粘弹性地壳变形分析，开发了一种数据驱动加速的可扩展非结构化隐式有限元求解器。在这里，我们结合了一个数据驱动的预测器，它使用过去的时间步长数据来估计高精度的初始解，以及一个基于多网格的共轭梯度求解器来有效地求解剩余的误差。与在块jacobi预条件共轭梯度解算器上使用标准初始解预测器相比，使用数据驱动预测器获得了3.19倍的加速，并且与多网格解算器结合使用，在Fugaku上获得了76.8倍的总加速。此外，由于数据驱动预测器的计算是局部化的，可以在计算节点之间进行通信，求解器在Fugaku的73728个计算节点上获得了78.5%的高弱可扩展性效率，使得整个应用的FP64峰值效率达到6.88%。这种发展也有望有助于加速其他基于pde的时间演化问题。
Targeting viscoelastic crustal deformation analysis, we develop a scalable unstructured implicit finite-element solver accelerated by a data-driven method. Here, we combine a datadriven predictor, that uses past time step data for estimating high-accuracy initial solutions, and a multi-grid based conjugate gradient solver for efficient solving of the remaining errors. When compared to using a standard initial solution predictor on a block Jacobi-preconditioned conjugate gradient solver, a 3.19fold speedup was attained by using the data-driven predictor, and combination with a multi-grid solver attained a total speedup of 76.8-fold on Fugaku. Furthermore, as the computation of the data-driven predictor is localized and can be conducted without communication between computation nodes, the solver attained high weak scalability efficiency of 78.5% up to 73728 compute nodes of Fugaku, leading to 6.88% FP64 peak efficiency for the whole application. Such development is also expected to be useful for accelerating other PDE-based time evolution problems.