CPU-GPU hybrid parallel binomial American option pricing

Nan Zhang*, Eng Gee Lim, Ka Lok Man, Chi Un Lei

*Corresponding author for this work

Research output: Chapter in Book or Report/Conference proceedingConference Proceedingpeer-review

6 Citations (Scopus)

Abstract

We present in this paper a novel parallel binomial algorithm that computes the price of an American option. The algorithm partitions a binomial tree constructed for the pricing into blocks of multiple levels of nodes, and assigns each such block to multiple processors. Each of the processors then computes the option's values at its assigned nodes in two phases. The algorithm is implemented and tested on a heterogeneous system consisting of an Intel multi-core processor and a NVIDIA GPU. The whole task is split and divided over and the CPU and GPU so that the computations are performed on the two processors simultaneously. In the hybrid processing, the GPU is always assigned the last part of a block, and makes use of a couple of buffers in the on-chip shared memory to reduce the number of accesses to the off-chip device memory. The performance of the hybrid processing is compared with an optimised CPU serial code, a CPU parallel implementation and a GPU standalone program.

Original languageEnglish
Title of host publicationInternational MultiConference of Engineers and Computer Scientists, IMECS 2012
PublisherNewswood Limited
Pages1157-1162
Number of pages6
ISBN (Print)9789881925190
Publication statusPublished - 2012
Event2012 International MultiConference of Engineers and Computer Scientists, IMECS 2012 - Kowloon, Hong Kong
Duration: 14 Mar 201216 Mar 2012

Publication series

NameLecture Notes in Engineering and Computer Science
Volume2196
ISSN (Print)2078-0958

Conference

Conference2012 International MultiConference of Engineers and Computer Scientists, IMECS 2012
Country/TerritoryHong Kong
CityKowloon
Period14/03/1216/03/12

Keywords

  • Binomial method
  • Graphics processing unit
  • Heterogeneous processing
  • Option pricing
  • Parallel computing

Fingerprint

Dive into the research topics of 'CPU-GPU hybrid parallel binomial American option pricing'. Together they form a unique fingerprint.

Cite this