Boston QWAFAFEW Meeting: Tuesday, 20 Nov 2012

Scaling MATLAB Financial Applications with Clusters, GPUs and the Cloud’

Advances in parallel and distributed computing technologies and reduction in the cost of computational hardware has opened up many opportunities for quants who rely on intensive computations for their research. Gone are the days when quants had to migrate their applications written in high-level languages like MATLAB to low-level languages like C or C++ in pursuit of performance gains. In this talk, we will discuss and demonstrate how MATLAB applications can be scaled with parallel computing technologies, Graphical Processor Units (GPUs), web technologies and the Amazon EC2 cloud. We will highlight the kinds of applications suited for parallel and distributed computing methodologies through examples and case studies. We will make a case on how quants can leverage these new technologies and obtain performance gains without having to significantly alter their research process.