Cost vs. performance of VaR on accelerator platforms

Authors:
Mukul Majmudar;Ciprian Docan;Manish Parashar;Christopher Marty
Affiliations:
Rutgers University, Piscataway, NJ;Rutgers University, Piscataway, NJ;Rutgers University, Piscataway, NJ;Bloomberg LP, New York, NY
Venue:
Proceedings of the 2nd Workshop on High Performance Computational Finance
Year:
2009

Citing 4
Cited 1

High dimensional pricing of exotic European contracts on a GPU Cluster, and comparison to a CPU cluster

IPDPS '09 Proceedings of the 2009 IEEE International Symposium on Parallel&Distributed Processing
Advanced risk analytics on the cell broadband engine

IPDPS '09 Proceedings of the 2009 IEEE International Symposium on Parallel&Distributed Processing
Option pricing with COS method on graphics processing units

IPDPS '09 Proceedings of the 2009 IEEE International Symposium on Parallel&Distributed Processing
On the energy efficiency of graphics processing units for scientific computing

IPDPS '09 Proceedings of the 2009 IEEE International Symposium on Parallel&Distributed Processing

Finding the right level of abstraction for minimizing operational expenditure

Proceedings of the fourth workshop on High performance computational finance

Quantified Score

Hi-index	0.00

Visualization

Abstract

The computation of value at risk (VaR) can be parallelized to boost performance, but different parallel platforms entail different gains in performance, as well as different costs. This paper explores the cost and performance tradeoffs inherent in the computation of VaR when implemented on different parallel platforms.