Speaker
Description
Abstract: GPUs are widely used in many areas of science, scientific computing and real-time data processing. In this talk, I will briefly introduce some of GPUs' uses in the (near) real-time data processing in radio astronomy for the Square Kilometre Array (SKA) radio telescope. The Square Kilometre Array (SKA) is an international initiative for developing the world's largest radio telescope with a total collecting area of over a million square meters. The data processing in radio astronomy is for the most part low in arithmetical intensity which necessitated good data reuse in the GPU implementations, and it has to deal with an unknown of inputs or number of samples on vastly different scales. A particular example of this is when an image is reconstructed from frequency-domain data (gridding) and when an image is transformed back to frequency domain (de-gridding) in the iterative image cleaning process.
