Presenter

Mark J. Stock is a performance engineer at Hewlett Packard Enterprise where he helps scientists optimize their codes for running on leadership-class supercomputers. His two decades of experience in parallel computing and computational physics support this work, which requires understanding the layers of an application, from GPU kernels through parallel algorithms to schedulers and network tuning. He holds a PhD in aerospace engineering from the University of Michigan. His interests include computational fluid dynamics, GPU computing, and generative artwork. Mark lives and works in Boston, Massachusetts, USA.