This is a repository of pywren examples, showing how to run various
example code and generate many of the plots used in blog posts.
Most examples have an explanation in a
README.md, a script to run, and
often a Jupyter notebook for interactively examining
Note that these examples, in addition to requiring the latest pywren, often require additional packages like Jupyter/iPython, Matplotlib, Seaborn, and the Ruffus pipeline manager.
Hello world is a simple example to get you up and running with pywren.
TFLOPS on microservices
An example of how to achieve over 40 TFLOPS of numerical performance using pure-Python code running on thousands of simultaneous cores. This example is based on our original blog post and our recent paper. [[code]]flops_benchmark)
GB/s from S3
We can achieve up to 80 GB/sec read and 60 GB/sec write performance to S3 in this benchmark example, based on our original blog post. We have notebooks that show how to benchmark and then how to measure scaling. [code].
Measuring Lambda's recycling
Running a parameter server
Robust Kalman Filtering
Inverse problems with sweep