Tapping the supercomputer under your desk: Solving dynamic equilibrium models with graphics processors

被引：32

作者：

Aldrich, Eric M. ^{[1
,4
]}

Fernandez-Villaverde, Jesus ^{[2
,3
]}

Gallant, A. Ronald ^{[4
,5
]}

Rubio-Ramirez, Juan F. ^{[1
,4
]}

机构：

[1] Fed Reserve Bank Atlanta, Atlanta, GA 30030 USA

[2] Univ Penn, Philadelphia, PA 19104 USA

[3] NBER, Cambridge, MA 02138 USA

[4] Duke Univ, Durham, NC 27706 USA

[5] NYU, New York, NY 10003 USA

来源：

JOURNAL OF ECONOMIC DYNAMICS & CONTROL | 2011年 / 35卷 / 03期

关键词：

CUDA; Dynamic programming; Parallelization; Growth model; Business cycles;

D O I：

10.1016/j.jedc.2010.10.001

中图分类号：

F [经济];

学科分类号：

02 ;

摘要：

This paper shows how to build algorithms that use graphics processing units (GPUs) installed in most modem computers to solve dynamic equilibrium models in economics. In particular, we rely on the compute unified device architecture (CUDA) of NVIDIA CPUs. We illustrate the power of the approach by solving a simple real business cycle model with value function iteration. We document improvements in speed of around 200 times and suggest that even further gains are likely. (C) 2010 Elsevier B.V. All rights reserved.

引用

页码：386 / 393

页数：8

共 5 条

[1] Comparing solution methods for dynamic equilibrium economies [J].

Aruoba, S. Boragan ;

Fernandez-Villaverde, Jesus ;

Rubio-Ramirez, Juan F. .

JOURNAL OF ECONOMIC DYNAMICS & CONTROL, 2006, 30 (12) :2477-2508

[2]

Heer Burkhard., 2005, Dynamic General Equilibrium Modelling

[3] PROJECTION METHODS FOR SOLVING AGGREGATE GROWTH-MODELS [J].

JUDD, KL .

JOURNAL OF ECONOMIC THEORY, 1992, 58 (02) :410-452

[4]

LEE A, 2009, UTILITY GRAPHI UNPUB

[5] FINITE STATE MARKOV-CHAIN APPROXIMATIONS TO UNIVARIATE AND VECTOR AUTOREGRESSIONS [J].

TAUCHEN, G .

ECONOMICS LETTERS, 1986, 20 (02) :177-181

← 1 →