John Urbanic
Pittsburgh Supercomputing Center
urbanic@psc.edu

Slides: Introduction to OpenACC, Advanced OpenACC, OpenACC with CUDA, OpenMP 4 Sneak Peek,Hybrid Programming, Exercises

After our Parallel Programming Methodologies overview you may have the impression that GPUs have become an incredibly important part of the modern computing landscape. Actually, you probably already have an idea that that might be true. What you may not know is how to take advantage of this capability. The good news is that the low-level approaches of the "early days" are no longer necessary to get efficient and flexible GPU usage. OpenACC is the well-supported standard that allows us to access the very different kind of computing capability that has become central to numerical computing, data analytics, and beyond.

John Urbanic is Parallel Computing Scientist at the Pittsburgh Supercomputing Center (PSC). At PSC, he spends as much time as possible implementing extremely scalable code on interesting machines.  These days that means using a lot of MPI, OpenMP and OpenACC. He has also been sucked into the world of Big Data challenges and Deep Learning. John teaches workshops and classes in his specialty areas and leads the XSEDE Monthly Workshop Series, Summer Boot Camp, in addition to the International HPC Summer School on HPC Challenges in Computational Sciences.  John has physics degrees from Carnegie Mellon University (BS) and Pennsylvania State University (MS) and still appreciates working on applications that simulate physical phenomena.

Last modified: Wednesday, June 28, 2017, 12:04 PM