High-Time Resolution GPU Imager for Low-Frequency Radio Telescopes

Abstract

This thesis describes an implementation of a high-time-resolution GPU imager for radio astronomy data, which would be a part of the larger GPU-based processing pipeline searching for FRBs and its benchmarks on Topaz and Setonix supercomputers. Finally, the thesis also showcases the results of a pilot FRB survey using a few hours of 100 ms data from the SKA-Low prototype station (EDA2).