Follow Datanami:
April 5, 2013

ASTRON, IBM to Help Researchers Listen to SKA

Ian Armas Foster

Understanding the cosmos and understanding our origins are one in the same. For the past few years, understanding big data has become a part of that process as telescopes are capable of retrieving a sizable amount of complex information.

The Square Kilometre Array (SKA) will seek to exponentially surpass current telescopes and thus delve into the formation of stars near the beginning of the known universe. To facilitate SKA’s immense computational needs, ASTRON, the Netherlands Institute for Radio Astronomy, and IBM have teamed up to form DOME in researching the computing side of the highly ambitious astronomical survey project. Last week, the two held a recruiting session for MS and PhD students heading toward a degree in the sciences, engineering, or IT to come to the newly formed joint research facility in the Dutch town of Dwingeloo.

The array, whose telescopes will cover an area roughly equal to that of the continental United States, would have to collect and process about an exabyte a day. For comparison’s sake, today’s internet also produces roughly an exabyte of data every day.

Obviously, processing such an amount today is slightly unreasonable. However, by the time SKA becomes fully operational in the year 2024, the hope is that ASTRON and IBM will have tackled these issues as computing power continues to grow over the next eleven years. The below video outlines ASTRON’s and IBM’s plans to attack the massive computing problem they have before them.

As shown above, data transfer, storage, streaming, and analytics will be among the technological challenges that IBM and ASTRON will face. Some ideas include three-dimensional chip stacking, which would allow for higher storage density. DOME is also considering photonic upgrades to optical fiber transmission systems for quicker and more efficient data movement.

All of these ideas are in their infancy as the international consortium builds toward the 2024 opening of the array to be placed in either Australia or South Africa.

A research project as ambitious and intriguing as SKA is a way to ensure that big data study and innovation doesn’t get stuck in just culling social and web information for the purposes of optimizing marketing campaigns. Further, a goal that lies eleven years away will give participants plenty of time to brainstorm and implement newer technologies in experimental scales such that they are properly evaluated before SKA’s switch is flipped.

Related Articles

IBM: Five Use Cases Driving Big Data Adoption

Big Science and the Gift that Keeps on Giving

What Can Enterprises Learn From Genome Sequencing?