Explore Python profiling techniques using Intel® VTune™ Amplifier in this 33-minute conference talk from EuroPython 2017. Delve into the growing significance of Python in high-performance computing and machine learning, and learn how to leverage a powerful GUI application to analyze runtime performance, identify bottlenecks, and visualize call-stacks. Discover various profiling types, examine code examples, and gain insights into CPU time analysis, timeline filtering, and mixed-mode analysis. Acquire practical knowledge on optimizing Python applications, including tips on using debug flags, running multiple iterations, and interpreting source line information.
Overview
Syllabus
Introduction
Python popularity
Performance analysis
Mathematics and data science
Intel Distribution
NKL
Vectorization
In the coming 1015 minutes
How do we find those bottlenecks
Why do we need a byte optimization
Python is everywhere
Web frameworks
Code examination
Logging
Profiling
Profiling Types
Overview
Distributions
Running Process
CPU Time
Call Stack
Open Source
Zoom in Timeline
Filter in Timeline
Shared Objects
Mixed Mode Analysis
Summary
More information
Annotation
Debug flag
Run multiple times
Source line
Interpreter
Workshop Question
Taught by
EuroPython Conference
