About black_scholes_example:

Options are becoming an increasingly important financial tool as derivative
financial instrument. Option pricing models such as Black-Scholes are the
fundamental building blocks of computational finance. In this example we
demonstrate how easily one can program and speed up a Black-Scholes
computation using Jacket within MATLAB. This algorithm has a computation
pattern that has high arithmetic intensity and extensive parallelism making it
a good candidate for computation on the GPU. Notably, it uses transcendental
functions heavily, such as log, exp, sqrt etc. We demonstrate the
effectiveness of Jacket by using a simple cast of input data to Jacket's GPU
data types to speedup this code without any other modifications.



Instructions to run:

1) Add path to Jacket's engine directory in MATLAB (on Windows, running
StartJacket.bat will do this for you) and run black_scholes_example.m by
typing 'black_scholes_example' on the MATLAB prompt.

2) To specify the precision of the data to be used, please run the example
using black_scholes_example('single') or black_scholes_example('double') as
needed. If the card in use is capable of double precision arithmetic, the
double version is run by default.
