Time Propagation of Partial Differential Equations Using the Short Iterative Lanczos Method and Finite-Element Discrete Variable Representation: An Experiment Using the Intel Phi Coprocessors

Published

July 15, 2016

Author(s)

Barry I. Schneider, klaus Bartschat, Xiaoxu Guan

Abstract

The Short Iterative Lanczos (SIL) method has been combined with the Finite-Element Discrete Variable Representation (SIL-FEDVR) to yield a powerful approach to solving the time-dependent Schroedinger equation. It has been applied to the interaction of short, intense laser radiation (attosecond pulses) to describe the single and double ionization of atoms and molecules, but the approach is not limited to this particular application. The paper begins with a brief description of the method and algorithms and then discusses how they have been successfully ported to the Intel Phi coprocessors. While further experimentation is needed, the results provide reasonable evidence that by suitably modifying the code to combine MPI, OpenMP, and compiler assisted offload (CAO) directives, one can achieve a significant improvement in performance from these coprocessors for problems such as the above. Future work will examine porting the code to GPU's and a comparison in performance of the two coprocessors.

Proceedings Title

XSEDE 2016

Conference Dates

July 17-21, 2016

Conference Location

Miami, FL

Pub Type

Conferences

Download Paper

DOI Link

Keywords

time-dependent Schroedinger equation, FEDVR, SIL, Phi coprocessors, high performance computing, parsllel procssing

Citation

Schneider, B. , Bartschat, K. and Guan, X. (2016), Time Propagation of Partial Differential Equations Using the Short Iterative Lanczos Method and Finite-Element Discrete Variable Representation: An Experiment Using the Intel Phi Coprocessors, XSEDE 2016, Miami, FL, [online], https://doi.org/10.1145/2949550.2949565 (Accessed April 24, 2024)