Tutorial Details: GPU CUDA Matlab basic
|Date:||Friday, February 28, 2014, 01:00 pm - 03:00 pm|
|Instructor(s):||Ravishankar Chityala, MSI, Kai Mollerud,|
Matlab currently is one of the most practical software used for numerical calculations. This tutorial will introduce the basic components of general purpose GPU (GPGPU) computing that can dramatically accelerate some Matlab applications. In addition to accessing the hardware resources (Cascade cluster, Koronis and GPUT nodes), we will discuss what type of applications can be run, how to execute the calculations on GPU device, and what kind of performance gains it can provide.
A hands-on practice session will follow the lecture. Provided examples will enable the attendees to get direct experience of porting a matlab code and executing it on GPU and CPU, so that one can compare the performance and evaluate its enhancement by GPU.