The Square Kilometre Array (SKA) project is an effort to build the world’s largest radio telescope, with a collecting area of over one square kilometre. The design and development of the SKA is a truly global effort involving 100 organizations across 20 countries. The SKA is one of the largest scientific endeavors in history and its scale represents a huge leap forward in both engineering and research & development.
The transformational science that can be conducted with the SKA includes challenging Einstein’s seminal theory of relativity, looking at how the very first stars and galaxies were formed, helping scientists understand dark energy, and answering questions about other life in the universe.
The SKA project relies on NVIDIA Tesla GPUs to process over 2,000 observing beams, which produce around 160GB/s of data, equivalent to around 50 hours of HD television data per second. In the video below, Wes Armour, Director at Oxford eResearch Centre, describes the role of GPUs in processing large amounts of astronomical data collected by the SKA, the use of the cuFFT CUDA-X library and NVIDIA® Nsight™ tools in their work, and how CUDA is the best suited and most performant option for their signal processing software.
During his session at GTC 2019, Armour talked about AstroAccelerate, a GPU enabled software package that uses CUDA and NVIDIA GPUs to achieve real-time processing of radio-astronomy data. He stated that “The massive computational power of modern day GPUs allows code to perform algorithms such as de-dispersion, single pulse searching and Fourier Domain Acceleration Searching in real-time on very large data-sets which are comparable to those which will be produced by next generation radio-telescopes such as the SKA.”