Speakers: Sergey Maidanov, Intel and Stan Seibert, Continuum Analytics
This session provides an overview of the tools and optimizations that Intel brings to Python* developers. High-performance libraries and profilers and extended support of parallelism at all levels achieve near-native performance in Python, avoiding the need to rewrite in C or C++ to put a project in production. Our case studies show up to 100 times the speedups up in NumPy, SciPy*, PyDAAL, and scikit-learn* and scaling across multiple cores and nodes. Learn how Intel® VTune™ Amplifier allows low overhead profiling of Python and native codes to identify performance hotspots. Obtain near-native code performance with tools like Cython and Numba*.