Analysis
NVIDIA GPU-Accelerated Supercomputer Sets World Record For Energy Efficiency
NVIDIA have announced that Italy’s “Eurora” supercomputer – which uses NVIDIA Tesla GPU accelerators based on NVIDIA Kepler, the world’s fastest and most efficient high performance computing architecture – has set a new record for data center energy efficiency.
The Eurora achieved the record-breaking achievement by combining 128 high-performance, energy-efficient NVIDIA Tesla K20 accelerators with the Eurotech Aurora Tigon supercomputer, featuring innovative Aurora Hot Water Cooling technology, which uses direct hot water cooling on all electronic and electrical components of the HPC system.
Available to members of the Partnership for Advanced Computing in Europe (PRACE) and major Italian research entities, Eurora will enable scientists to advance research and discovery across a range of scientific disciplines, including material science, astrophysics, life sciences and Earth sciences.
“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.”
Pairing NVIDIA Tesla K20 GPUs with Eurotech’s Aurora Hot Water Cooling technology, the Eurora system is more efficient and compact than conventional air-cooled solutions. HPC systems based on the Eurora hardware architecture, including the Eurotech Aurora Tigon, enable data centers to potentially reduce energy bills by up to 50 percent and reduce total cost of ownership by 30-50 percent.
In addition, the use of Aurora Hot Water Cooling technology reduces or eliminates the need for air conditioning in typically warm climates like Italy. The thermal energy the system produces can be used to heat buildings, drive adsorption chillers for air conditioning or generate tri-generation, the combined production of electricity, heating and cooling.
“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.”
Eurora is a prototype system developed for Cineca under the PRACE 2IP initiative to provide a sustainable, high-quality infrastructure to meet the most demanding needs of the European HPC user community. A commercial version of the Eurotech Aurora Tigon supercomputer is also available today from Eurotech.
NVIDIA Tesla GPUs are massively parallel accelerators based on the NVIDIA CUDA parallel computing platform and programming model. Tesla GPUs are designed from the ground up for power-efficient, high performance computing, computational science and supercomputing, delivering dramatically higher application acceleration for a range of scientific and commercial applications than a CPU-only approach.