In this webinar, you’ll see how MATLAB supports CUDA kernel development by providing a high level language and development environment for prototyping algorithms and incrementally developing and testing CUDA kernels.  Product demonstrations will highlight how MATLAB can be used to:

• Write prototype code to explore algorithms before implementing them in CUDA
• Quickly evaluate kernels for different input data
• Analyze and visualize kernel results
• Write test harnesses to validate that kernels are working correctly

You will see how MATLAB reduces the amount of code required for evaluating and testing kernels compared with lower level languages such as C or Fortran.  You will also see how the GPU-enabled functionality in MATLAB lets you take advantage of GPU computing without having to write CUDA kernels or learn low-level GPU computing libraries.

Previous knowledge of MATLAB is not required for this webinar.

About the Presenters:

Dan Doherty, MathWorks
Dan works as a Partner Manager at MathWorks, focusing on NVIDIA and other partners in the HPC area. Prior to working as Partner Manager, Dan was a Product Manager at MathWorks for over 5 years, focusing on MATLAB and core math and data analysis products.  Dan received a B.S.E. and M.S.E. in Mechanical Engineering from the University of New Hampshire, where his research focused on prediction of cutting forces during CNC machining.

Jonathan Bentz, NVIDIA
Jonathan Bentz is a Solution Architect with NVIDIA, focusing on Higher Education and Research customers. Prior to NVIDIA Jonathan worked for Cray as a software engineer in the Scientific Libraries group working on dense linear algebra and FFT software. Jonathan obtained a PhD in physical chemistry and an MS in computer science from Iowa State University.