The new hardware at the University of Cambridge will help process radio signals created during the Big Bang
In time, this high performance infrastructure will be used to analyse the output of the world’s largest radio telescope, the Square Kilometre Array (SKA), currently under construction in Africa and Australia.
The supercomputer is also one of the world’s greenest, using fresh air to efficiently cool the servers – although others, including IBM and HP, would wargue that liquid cooling is likely to be ultimately the most efficient solution in the long term.
“We’re thrilled to be working on one of the most important science projects of this decade and the experience and technology we’ve now built will go a long way in supporting future projects. We’ve received international acknowledgment for the HPC cluster and the energy efficiency of the system has kept up with our internal goals,” said Paul Calleja, director of HPC Services at the University of Cambridge.
The edge of the universe
The SKA is an extremely powerful survey telescope which will use millions of antennas to gather radio signals, forming a collection area equivalent to one square kilometre, but spanning a huge surface area – over 3000 kilometres (1864 miles) wide. It will be 50 times more sensitive than any previous radio device and more than 10,000 times faster than today’s instruments.
Astronomers will use this power to study radio signals that were produced more than 13 billion years ago during the Big Bang. The collected signals are expected to generate one exabyte of raw data every day.
Prototype SKA systems will be deployed in 2016, and the construction of Phase 1 will take place from 2018 to 2023.
In order to participate in the project, Dell and its partners, Nvidia and Mellanox, had to design an energy-efficient High Performance Computing (HPC) cluster for the University of Cambridge that could handle extremely large amounts of data. For this purpose, the company deployed 128 Dell PowerEdge T620 servers with Xeon processors, for a total computational performance of 240 Teraflops.
The resulting supercomputer ranks second in the Green500 index, making it the most efficient air-cooled HPC cluster in the world.
“Dell’s Fresh Air Cooling was an ideal fit for our needs,” explained Calleja. “We needed to create the most efficient supercomputer in the world in order to back the prestige of the SKA project and we couldn’t do that with energy draining technology. Dell’s server solutions let us increase compute power while decreasing our energy expenditure.”
What do you know about supercomputers? Take our quiz!