Shuffle up and deal: accelerating GPGPU Monte Carlo simulation with application to option pricing

Summary In this paper, we demonstrate some speedup opportunity regarding Monte Carlo simulation on graphic processing unit architecture, with financial application. We leverage on the possibility of reducing the volume of actually generated random numbers, by replacing the generation phase with some...

Full description

Saved in:
Bibliographic Details
Published in:Concurrency and computation 2015-12, Vol.27 (17), p.5203-5213
Main Authors: Cassagnes, Aurelien, Chen, Yu, Ohashi, Hirotada
Format: Article
Language:English
Subjects:
Architecture
Computation
Computer simulation
Concurrency
CUDA
Error analysis
GPGPU
Graphics processing units
Monte Carlo
Monte Carlo methods
option pricing problem
Random numbers
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Summary In this paper, we demonstrate some speedup opportunity regarding Monte Carlo simulation on graphic processing unit architecture, with financial application. We leverage on the possibility of reducing the volume of actually generated random numbers, by replacing the generation phase with some shuffling using Compute Unified Device Architecture's built‐in shuffle instructions. We will study various shuffling patterns and duration, elect the best among them with regard to induced correlation, using Granger causality test. We will then study the accuracy and variance of results actually achieved by our general‐purpose computing on graphic processing unit shuffled Monte‐Carlo, exhibiting a computational time reduced by half while error remains marginal. Copyright © 2015 John Wiley & Sons, Ltd.
ISSN:1532-0626
1532-0634
DOI:10.1002/cpe.3561