Understanding the Potential of Interpreter-based
Optimizations for Python

The increasing popularity of scripting languages as general purpose programming environments calls for more efficient execution. Most of these languages, such as Python, Ruby, PHP, and JavaScript are interpreted. Interpretation is a natural implementation given the dynamic nature of these languages and interpreter portability has facilitated wide-spread use. In this work, we analyze the performance of CPython, a commonly used Python interpreter, to identify major sources of overhead. Based on our findings, we investigate the efficiency of a number of optimizations and explore the design options and trade-offs involved.