Testing Python 3.5's various concurrency methods

A little suite of tests for gauging the speed of various Python concurrency methods.

This suite tests:

• CPU Bound operations
• I/O Bound operations
• A mix of CPU and I/O operations

Each of the tests contains 5 parts:

• Sequential execution (for comparison).
• Multiprocessing pool execution (classical Python multiprocessing)
• Multithreading pool execution (classical Python threading)
• Python 3.5 async Coroutines (using a multiprocessing pool)
• Python 3.5 async Coroutines (using a threading pool)

Each of the 3 test types should be taken seperately as they contain different computations.

Test it yourself

git clone https://github.com/Sonictherocketman/python3.5_coro_testing
cd python3.5_coro_testing
bash run.sh

Sample Results

Retina Macbook Pro (2014, 13")

<tr>
    <td>I/O Bound</td>
    <td>Serial</td>
    <td>1.924s</td>
</tr>
<tr>
    <td></td>
    <td>Coro Multi</td>
    <td>0.951s</td>
</tr>
    <td></td>
    <td>Coro Thread</td>
    <td>1.97s</td>
</tr>
<tr>
    <td></td>
    <td>Multiprocess</td>
    <td>1.320s</td>
</tr>
<tr>
    <td></td>
    <td>Threading</td>
    <td>1.807s</td>
</tr>

<tr>
    <td>Mixed I/O and CPU</td>
    <td>Serial</td>
    <td>4.789s</td>
</tr>
<tr>
    <td></td>
    <td>Coro Multi</td>
    <td>13.574s</td>
</tr>
    <td></td>
    <td>Coro Thread</td>
    <td>7.724s</td>
</tr>
<tr>
    <td></td>
    <td>Multiprocess</td>
    <td>4.513s</td>
</tr>
<tr>
    <td></td>
    <td>Threading</td>
    <td>4.976s</td>
</tr>