This live webinar gave an overview of the NAG PyPOP tool and how it can be used to streamline the process of application performance analysis. Phil Tooley demonstrated the PyPOP analysis workflow, showed how the PyPOP wizard can be used to quickly calculate the POP metrics from captured profiling data and generate a performance report. He also introduced some of the more advanced features such as the OpenMP region explorer and showed how PyPOP can be easily extended to perform customised analyses.
The presentation slides are also available here.
About the Presenter
Phil Tooley is one of our POP performance experts and the lead author of the PyPOP analysis tool. He has a wide range of experience in both academic and industrial HPC projects both on-premises and in the cloud and is passionate about helping people write fast, efficient software.
Phil began his career studying computational plasma physics, where he developed large-scale HPC applications using MPI and OpenMP. He quickly discovered that writing great code was his passion and moved fully into the world of high performance computing. Along with traditional HPC development, Phil also champions the use of modern HPC paradigms such as Spark and Flink, along with high-performance Python using Numpy+Numba for the development of modern HPC applications.