Computer Science

Parallel Computation of Sovereign Default Models

Document Type

Article

Abstract

This paper discusses the parallel and efficient computation of macroeconomic models, with an emphasis on solving sovereign default models. Our motivation is twofold. First, we aim to streamline complex numerical models in a parallel computation fashion. Second, we want to unleash the power of graphic processing unit but bypass the steep learning and implementation costs of languages like C++ and Compute Unified Device Architecture (CUDA) in economic research. To this end, we propose a framework for efficient parallel computing with the modern language Julia. The paper offers a detailed analysis of parallel computing, Julia-style acceleration techniques, and coding recommendations. The Julia with CUDA benchmark shows a substantial speedup of over 1000 times compared to standard Julia that runs on a CPU. Our Julia with CUDA’s implementation is twice as fast as that of the C++ Standard Parallel Library. We provide an accompanying GitHub repository with the codes and the benchmarks used in this paper.

Publication Title

Computational Economics

Publication Date

2023

Volume

62

Issue

3

First Page

1047

Last Page

1085

ISSN

0927-7099

DOI

10.1007/s10614-022-10291-1

Keywords

C++, CUDA, GPU, Julia, parallel computation, sovereign default model

APA Citation

Deng, M., Guerron-Quintana, P. A., & Tseng, L. (2023). Parallel Computation of Sovereign Default Models. Computational Economics, 62(3), 1047-1085.

Link to Full Text

Share

COinS