Saving processing time by processing radio astronomy data on multiple machines

What is it about?

To make images, LOFAR data needs to pass through a direction independent (DI) calibration pipeline. This processing can be parallelized by being split and run concurrently on many machines. This work describes a method to easily isolate and process parts of the full data. This makes the processing of LOFAR data faster, and easier to automate. Finally, we can scale this processing across cloud infrastructure making it easy to handle large amounts of data.

Featured Image