This webinar is intended to help those who are just starting in High Performance Computing (HPC) get acquainted with the basics of HPC allocations as well as job submissions and profiling. We will begin with explaining the basics of obtaining an allocation to a supercomputer through the XSEDE program. Next, we will describe the differences between the login nodes and the compute nodes, and the choices of schedulers (e.g. PBS, TORQUE, SLURM). Finally, we will show how to submit a job to a supercomputer and check the status of the job, as well as demonstrate some profiling tools like top to verify that the job is running as expected. Please note that while we will work with Stampede for the purposes of this webinar, the same principles can be applied to other supercomputers. A simple OpenMP and MPI program will be used for the demonstration. No previous programming experience is required.
Biography: Pedro David Bello-Maldonado is a PhD student in Computer Science at the University of Illinois at Urbana-Champaign (UIUC) since 2014 and a former XSEDE scholar from 2013-2015. Before coming to Urbana-Champaign, Pedro completed extensive coursework in Mechatronics Engineering and Electrical Engineering in his native country, Colombia, and then later in the United States. He received his BS and MS in Electrical Engineering from Florida International University in 2013 and 2014 respectively, where he was also honored to be awarded the Outstanding Graduate Student Award in recognition of his academic achievements. Pedro brings his unique multicultural background and his natural curiosity to his current research at UIUC which focuses on the solution of partial differential equations using numerical methods and high performance computing in computational fluid dynamics applications.
This webinar was assisted by David Manosalvas, a PhD Student in Aeronautics and Astronautics at Stanford University, and a former XSEDE Scholar from 2012-2015.