Myths and fairy tales around Python performance

Track:
Python Core, Internals, Extensions
Type:
Talk (long session)
Level:
intermediate
Duration:
45 minutes

Abstract

Python is slow: why? Can we make it faster?

Over the years, various answers to these questions have been given, and in my opinion, many of those are partial, imperfect, or just plainly wrong.

The truth is that there is no simple answer. We will examine some of the most common ones, and explain why they aren’t totally accurate. While doing so, we will examine the current status of some of the techniques currently adopted such as static typing, JIT, and AOT compilation, and explain why those alone are not enough.

Finally, we will go deeper and try to understand what are the fundamental issues to overcome, and what could be possible ways of moving forward.