10 November 2017
Driving information technology for the world’s biggest radio telescope
SKA antennae positioned in the desert. SKA is due to come online in the early 2020s, and will allow astronomers to explore the universe in unparalleled detail.
Few undertakings are as ambitious and visionary as the Square Kilometre Array (SKA) project, which seeks to build the world’s biggest radio telescope. And at the very centre of this impressive venture is STFC supercomputing expertise.
Rather than a single machine, SKA will be made up of an array of different antennas spanning Australia, and South Africa. Together, these antennas will detect radio wave signals emitted by cosmic objects, including those that are very far away. This will allow scientists to ‘see’ previously unexplorable parts of space, and to study the history of the universe.
SKA is significantly more sensitive and faster than anything seen before, and the project will generate huge amounts of data. When it comes online in the early 2020s, SKA will allow astronomers to explore the universe in unparalleled detail.
But SKA isn’t just a huge leap forwards for science, it’s also an incredible feat of information technology. The project is thought to be the world’s largest public science data project, and is expected to generate several times more data than global internet traffic. Because of the vast amounts of information gathered by the various SKA antennas, cutting-edge supercomputing is instrumental to the project’s success.
The crucial computing element of SKA is called the Science Data Processor (SDP). As the telescope’s central computer, the SDP will exploit the most advanced supercomputing available today. The processor will ingest and refine data, providing it as a neat package for astronomers to interpret.
The SDP Consortium is charged with making all this happen. Construction for SKA begins in 2019, and so the design and planning for the processor is already well underway. Led by Professor Paul Alexander from Cambridge, the consortium has been working with more than 40 organisations around the world.
STFC is a key contributor toward efforts to bring the SDP to life, and both the Scientific Computing Department & RAL Space have contributed to the processor design. Expertise from these departments will help guide development of hardware platforms, software, and algorithms that will allow SDP to process and refine data.
RAL Space are working alongside the Universities of Oxford and Cambridge to test the design of the SDP. This is a crucial stage in ensuring the day one success of the technology. The group are tasked with checking the assumptions used in the model of the design against current and anticipated supercomputing standards. This process of benchmarking will help to inform important decisions about hardware and software.
Meanwhile, the STFC Hartree Centre is advancing power-efficient computing software that can handle the huge amounts of data created by large experimental research initiatives like SKA. STFC will continue to design elements of the software pipeline with help from astrophysicists at the University of Manchester.
Supercomputing is at the very core of SKA, and the design of the Science Data Processor will be fundamental to the success of this ambitious project. There’s much to consider when designing information technology for a mega-science project such as this, but STFC experts are up to the challenge.