Evaluating Python Implementations
This section presents the results of the evaluation (runtime performance and memory consumption) represented with figures and tables. You may want alternatively get the raw data or download the benchmark suite and execute it in your computer. The specific measurements can be known placing the mouse over the figure. The tables below the figures show the specific data. It can be ordered by clicking over the column title. Use the dropdown lists to add / delete / filter the data to be shown.
Performance of implementations, grouped by the type of benchmark
This section allows showing execution time data stacked by micro-benchmarks, benchmarks, programs and large-scale applications. By using the dropdown lists, the figure and table is filtered by code category (common, multi-threading, IO and metaprogramming), methodology (start-up and steady-state), and language (Python 2 and Python 3).
Performance of Python 2 implementations
Here, runtime performance of Python 2 implementations can be compared, adding and removing code categories (common, multi-threading, IO and metaprogramming) and kind of applications (micro-benchmarks, benchmarks, programs, and large-scale applications), using the dropdown lists. The methodology dropdown list allows filtering by start-up and steady-state. Data in the table below changes accordingly.
Performance of Python 3 implementations
In this section you can compare the runtime performance Python 3 implementations. Using the dropdown lists, you can add and remove code categories (common, multi-threading, IO and metaprogramming) and kind of applications (micro-benchmarks, benchmarks, programs, and large-scale applications). The methodology dropdown list allows filtering by start-up and steady-state. Data in the table below changes accordingly.
Memory consumption
Finally, we present the memory consumption relative to CPython 2. Both Python 2 and Python 3 can be selected and compared. Data in the table on the right changes accordingly.