Parallel Programming and Optimization
for High-Performance Computing (HPC)
2nd Floor, Room 202
12th Street & Montgomery Avenue
July 7 until July 11, 2014
The HPC team of the College of Science and Technology at Temple University with support of Temple’s Computer Services division and the Institute of Computational Molecular Science is conducting a one week long school on code optimization and parallel programming for HPC. The activity is offered at no cost for participants. Participants are requested to bring a laptop to be able to follow the hands-on exercises and access the HPC cluster.
This introductory activity will be suitable for computational scientists at the student, postdoctoral, faculty or staff level with little or no previous HPC experience except basic knowledge of programming in either C/C++ or Fortran or both. The school will particularly benefit participants who work in computationally intense research areas in physics, chemistry, biology, engineering, mathematics.
Through a combination of introductory lectures and intense hands-on training, the school focuses on progressively teaching participants fundamental skills to improve the performance of their scientific codes. Participants will achieve this by learning how to improve the utilization of modern HPC hardware on multiple levels. The workshop will begin with low-effort code optimization techniques, and then advance to distributed memory parallel programming and multi-threading tutorials. Participants will develop and test code exercises on Temple’s production HPC cluster “Owl’s Nest”. The Owl’s Nest software environment is modelled around those found on large scale national supercomputing facilities, thus giving participants valuable and transferable HPC skills.