Solving the incomplete markets model in parallel using GPU computing and the Krusell–Smith algorithm
Solving the incomplete markets model in parallel using GPU computing and the Krusell–Smith algorithm
This paper demonstrates the potential of graphics processing units in solving the incomplete markets model in parallel using the Krusell–Smith algorithm. We illustrate the power of this approach using the same exercise as in Den Haan et al. (J Econ Dyn Control 34:1–3, 2010). We document a speed gain which increases sharply with the number of agents. To reduce entry barriers, we explain our methodology and provide some example algorithms.
GPU computing, heterogeneous agents, incomplete markets, interpolation, krusell–smith algorithm
569-591
Hatcher, Michael
e0846252-6d46-44f8-ba3c-05cf1fba64ab
Scheffel, Eric M.
829e5ee1-d9f0-445e-aecb-2a99bcc118a6
December 2016
Hatcher, Michael
e0846252-6d46-44f8-ba3c-05cf1fba64ab
Scheffel, Eric M.
829e5ee1-d9f0-445e-aecb-2a99bcc118a6
Hatcher, Michael and Scheffel, Eric M.
(2016)
Solving the incomplete markets model in parallel using GPU computing and the Krusell–Smith algorithm.
Computational Economics, 48 (4), .
(doi:10.1007/s10614-015-9537-0).
Abstract
This paper demonstrates the potential of graphics processing units in solving the incomplete markets model in parallel using the Krusell–Smith algorithm. We illustrate the power of this approach using the same exercise as in Den Haan et al. (J Econ Dyn Control 34:1–3, 2010). We document a speed gain which increases sharply with the number of agents. To reduce entry barriers, we explain our methodology and provide some example algorithms.
Text
paper_mch_final_a.pdf
- Accepted Manuscript
More information
Accepted/In Press date: 23 November 2015
e-pub ahead of print date: 1 December 2015
Published date: December 2016
Keywords:
GPU computing, heterogeneous agents, incomplete markets, interpolation, krusell–smith algorithm
Organisations:
Economics
Identifiers
Local EPrints ID: 384832
URI: http://eprints.soton.ac.uk/id/eprint/384832
PURE UUID: 476b48f8-a510-44a4-8b93-5ac6268e8035
Catalogue record
Date deposited: 11 Jan 2016 12:05
Last modified: 15 Mar 2024 03:50
Export record
Altmetrics
Contributors
Author:
Eric M. Scheffel
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics
