I develop parallel algorithms for scientific applications at IBM Research - together we are accelerating science for emerging computing architectures for the new generation of massively-parallel high-performance computers. This is an extremely interesting place to be - it not only lets me think about mathematics and physics, but also allows me to apply my passion for computing to solve these problems.
When I’m not by a computer, I enjoy cooking, gardening, and exercising outdoors.
Growing up in a family with other physicists I naturally had a desire to learn more about the trade. However, I was fascinated by computers and had a bad habit of taking apart all the family electronics. I owe a lot to Peter Honeyman for introducing me formally to the world of system development. Through this experience I found a friendly and active open source community that I have always tried to stay a part of.
My primary graduate work was in the field of high-order computational fluid dynamics (CFD) for aerospace applications. My advisor Krzysztof Fidkowski and I developed a new class of discontinuous Galerkin (DG) methods that are computationally less expensive than existing methods while retaining the nice properties of DG iself. These methods transfer more of the computation to local element-wise operations which increase the parallelism, making them ideally suited for high-performance computing.