by Dan Doherty

NVIDIA Tesla and Quadro GPUs have the power to dramatically decrease compute time for your applications with MATLAB, a high-level computer language for performing computationally intensive tasks.

To help you leverage that power, MathWorks – which develops mathematical computing software for engineers and scientists – is offering a free webinar to get you on the path to GPU-accelerated computing.

A solution for a second-order wave equation on a
32 x 32 grid. Solving the equation on a GPU
can produce a 10x speed-up compared with a
CPU for large grid sizes.

The webinar is intended for MATLAB users who need to process large amounts of data or perform time-consuming simulations. MathWorks will hold their webinar on October 27, 2011 to teach MATLAB users how to make minimal changes to your existing code in order to perform computations using NVIDIA Tesla and Quadro GPUs – using massively parallel architecture to attain significant performance gains compared with CPUs.

Product demonstrations will highlight topics such as:

  • Using GPU enabled MATLAB functions
  • Using custom CUDA kernels from MATLAB
  • Distributing MATLAB GPU applications as standalone executables and software components

The webinar will last approximately 45 minutes, and will conclude with a Q&A session with a MathWorks engineer.

Register for the webinar here: www.mathworks.com/company/events/webinars/wbnr59816.html

You can also learn more by reading a recent newsletter article on GPU programming in MATLAB.