No matter how powerful the world's largest supercomputers are, some applications need even higher levels of performance. Research efforts around the world are developing hardware and software technologies to build massively parallel exascale computers, which are expected to provide unprecedented levels of application performance and acceptable power consumption. In the US, the Department of Energy (DOE) has launched the multi-billion dollar Exascale Computing Project (ECP), not only to tackle the challenge of developing capable exascale systems but also to create applications that are ready for them. In this talk, we discuss some parallel programming models that are part of ECP and the capabilities they are expected to offer. In particular, we describe the role of OpenMP in the anticipated exascale computing landscape.
//
Dr. Barbara Chapman
is a Professor of Applies Mathematics & Statistics and Computer Science at Stony Brook University. She is affiliated with the Institute for Advanced Computational Science (IACS) and has a joint appointment at the Brookhaven National Laboratory (BNL). She completed his Ph.D. at Queen's University of Belfast under the supervision of Professor Ronald Perrott. She was the director of the Center for Advanced Computing and Data Systems (CACDS) at the University of Houston. Professor Barbara Chapman has founded cOMPunity, Inc., an organization of researchers who are committed to furthering the work of the OpenMP Architecture Review Board (ARB), the organization that maintains and develops this application programming interface. Much of Dr. Chapman's current research is focused on OpenMP, a recent industry standard for shared memory parallel programming that has been broadly accepted by the computing community.