The next MCSP Seminar will be on Tuesday, March 26 at 11:30 in SC 134.
The featured speaker will be Adam Kucera, a 3Y Mathematics and Computer Science major and member of the Scholars Program. He will speak about his research experience at South Dakota State University last summer, which involved work on communication models for a supercomputer, specifically the development and implementation of various communication schemes related to Fast Fourier Transforms on a distributed processing system. His presentation is entitled, "Parallel Fast Fourier Transform." (See below for an
abstract.) As always, refreshments (cookies and beverages) will be provided. We hope to see you there!
March 26 Speaker: Adam Kucera, 3Y Mathematics and Computer Science major
"Parallel Fast Fourier Transform"
The Black-Scholes formula, one of the major breakthroughs in modern economics, allows for fast and easy computation of options pricing.
However, the basis of the formula, the Fast Fourier Transform, suffers from becoming too computationally heavy for serial computing models.
Developing models to compute the FFT portion of the Black-Scholes formula on a parallel architecture is very important to being able to calculate multiple option prices. To approach this we started with a generalized communication model for the FFT as well as attempting to implement this in a more standard computation library.
No comments:
Post a Comment