Eurora, a supercomputer located at the Cineca facility in Bologna, Italy, has set a new record for data centre energy efficiency.
The Eurotech-built machine generates 3,150 megaflops per watt of sustained performance, which is 26 per cent higher than the best system on the recent Green 500 supercomputer list.
The achievement has been attributed to Eurora’s high performance computing (HPC) architecture which features Nvidia’s Tesla K20 accelerators. The 128 GPUs are paired with Eurotech’s Aurora Tigon hot water cooling technology, which allows for direct hot water cooling on all the system’s electrical components.
The improvement in cooling efficiency could lead to a potential 50 per cent reduction in data centre energy bills, consequently slashing the overall cost of ownership by 30-50 per cent - the resulting thermal energy can also be used to heat buildings or as an alternative energy source to provide electricity for a company’s infrastructure.
“GPU accelerators are inherently more energy efficient than CPUs, and Tesla K20 accelerators widen this gap considerably,” said Sumit Gupta, general manger of the Tesla accelerated computing business at NVIDIA.
“Energy efficiency has become the defining element of computing performance. And GPUs enable data center computer systems of all sizes – from small clusters to future exascale-class systems – to achieve performance goals within an economically feasible energy budget,” he added
This record-setting piece of hardware will be used for research across a range of scientific disciplines and is available to all major Italian research groups as well as the Partnership for Advanced Computing in Europe (PRACE).
“Advanced computer simulations that enable scientists to discover new phenomena and test hypotheses require massive amounts of performance, which can consume a lot of power,” said Sanzio Bassini, director of HPC department at Cineca.
“Equipped with the ultra-efficient Aurora system and NVIDIA GPU accelerators, Eurora will give European researchers the computing muscle to study all types of physical and biological systems, while allowing us to keep data center power consumption and costs in check.”