Matt Bettencourt is currently a DevTech at NVIDIA, but is located in Bologna supporting the research conducted at the CINECA super computing center. He has over 25 years of HPC experience working through many different hardware paradigms, where he has developed applications in semiconductor physics, adaptive mesh refinement, climate weather and oceanography, and plasma physics all with a focus on HPC and extracting the most out of a particular hardware platform. Working with NVIDIA, Matt focuses on working with researchers to improve their code performance in a maintainable way through the use of standard programming models, abstraction layers, libraries and when needed, architecture specific programming models.