Abstract: OpenMP is the de facto standard for writing parallel applications for shared memory computers. Over 20 years of the development, with many new features and capabilities continuously being added, OpenMP has become a choice for performance professionals. However with a specification over 600 pages and growing, it may intimidate new OpenMP users and view the APIs as for "experts only".
Most OpenMP programmers, however, use just around 20 items from the specification. We call these items the “OpenMP Common Core”. By focusing on the common core first, we make OpenMP easy again for new users.
In this hands-on tutorial, we explore the common core of OpenMP through a carefully selected set of exercises, so students will master the common core and learn to apply it to their own problems. Please bring a laptop of your own with internet access. Training accounts will be provided for hands-on exercises.
Bio: Helen is a High Performance Computing Consultant at NERSC, Lawrence Berkeley National Laboratory. She serves as the main user focus point of contact, among users, systems and vendors staff, for the NERSC flagship systems including the Cray systems: XT4 (Franklin), XE6 (Hopper), and XC40 (Cori), deployed over the past 10 years, to ensure them are efficient for scientific productivity. She specializes in the software programming environment, parallel programming models such as MPI and OpenMP, applications porting and benchmarking, and climate models. Helen also serves on the OpenMP Language Committee to represent Berkeley Lab and has presented OpenMP tutorials at various venues including SC, XSEDE, IWOMP, and ECP. Helen has been on the Organizing Committee for many important HPC conference series, such as Cray User Group (Program Chair for 2017, 2018), SC, HPCS, IXPUG, and OpenMPCon. Helen has a Ph.D. in Marine Studies and an M.S in Computer Information Science.