The talk will outline benefits of Intel® Distribution for Python (IDP) for computationally intensive Python applications, as well as tools and techniques available to tune and optimize existing Python applications, including use of Intel® VTune™ Amplifier.
The Intel-NERSC joint collaboration on optimizing Python based project 'TomoPy' for Intel® Xeon™ and Intel® Xeon Phi™ for significant code speedups will be highlighted.
Bio: Oleksandr (Sasha) Pavlyk received his Ph.D. in Physics from Pennsylvania State University in 2005. He worked for Wolfram Research from 2005-2016 as a Mathematica Algorithms Developer, working on special functions, probability and statistics, etc. Pavlyk joined Intel Corp. in 2016 to work on Intel Distribution for Python, where he authored two packages, mkl_fft and mkl_random, providing Python users with access to respective Intel® MKL sub-components.