Parallel Black-Scholes Solver on Adaptive Sparse Grids

Sneha Jaiswal • Anushka Parmanand

Summary

We implement a parallel algorithm to solve the Black-Scholes equation for options pricing using the OpenMP programming model. Our approach parallelizes over a sparse grid with finite elements, which has a hierarchical tree structure and thus irregular memory access patterns. We also investigate the extent to which adaptive refinement, which improves pricing accuracy, degrades performance and examine how well dynamic scheduling can recover speedup.

Proposal

Download Proposal (PDF)

Milestone

Download Milestone (PDF)