Named after the nearby mountain peak, it is Europe’s leading supercomputer and is the third most powerful number cruncher in the world, after China’s Sunway TaihuLight and Tianhe 2 . (For those of us unfamiliar with these terms, a petaflop is the ability of a computer to do one quadrillion floating point operations per second, i.e. FLOPS).
The Piz Daint computer, run by CSCS, is located near Lugano by the Italian border. This supercomputer is used by Switzerland’s weather service for climate modeling and for research by scientists from around the world based in institutions which include the Institute of Particle Physics, the Human Brain Project, the universities of Basel, Bern and Genève and businesses such as the pharmaceutical company Novartis, insurance company Partner Re and Swiss Railways SBB.
How it is being used?
Using Piz Daint, scientists have had considerable success in analyzing the human genome, the origin of life on earth and extreme weather conditions. For example, with normal computers, searching for genes related to memory capacity has been comparable to seeking out the proverbial “needle in a haystack.” With the aid of the Piz Daint supercomputer, scientists at the University of Basel were able to discover interrelationships in the human genome that might simplify the search for “memory molecules” and eventually lead to more effective medical treatment for people with memory disturbance.
As for the origins of life on earth, the extreme temperature and pressure conditions of hydrothermal vents deep in the ocean, together with their distinct structures — a consequence of sulphide-rich minerals and the lack of light — are viewed as ideal breeding ground for chemical evolution of organic molecules, the building blocks of life. Researchers from Hungary, Switzerland and Italy are using Piz Daint to simulate the reaction chain that enables the formation of ammonia (NH3), which is one of the prebiotic molecules that are the key to the origin of life.
With regard to climate, Europeans are today witnessing with alarming regularity extreme weather events such as torrential rain on the northern side of the Alps and in central Europe, often causing the rivers Elbe, Danube and Rhine to flood. University of Bern scientists used climate modeling run on the “Piz Daint” to investigate the source of these abnormal rainfalls. They concluded that it was due to ‘volley ball’ cyclones picking up moisture from the warming surface of the Mediterranean Sea.
Some facts about Piz Daint
The original Piz Daint Cray XC30 system was installed in December 2012. This system was extended with Piz Dora, a Cray XC40 with 1256 compute nodes, in 2013. In October 2016, at a cost of 35 million Euros, Piz Daint and Piz Dora were upgraded and combined into the current XC50/XC40 system featuring NVidia Tesla P100 GPUs. According to the TOP500 website, the new system is the “second most energy-efficient supercomputer” on their list, behind an NVidia system that also uses P100 GPUs.
The supercomputers are located in a machine room, measuring some 2000 square meters with no single supporting pillar or any partitioning; this all to minimize restrictions to the installation and operation of supercomputers in the future.
CSCS has a very international working environment, with around 90 employees representing over 16 different nationalities, supervising over 1000 users conducting around 91 research projects.
The CSCS computing center uses as much electricity every day as a small town. About a third of this electricity is used for cooling. Computers get hot and must be cooled otherwise they melt. Piz Daint is cooled with up to 760 liters of water per second from nearby Lake Lugano. Using cool water from the lake significantly reduces overall electricity consumption. The water, taken 45 meters down is around 6 degrees. For ecological reasons, the water returning to the lake must never be over 25 degrees.
One thing is clear, the Piz Daint supercomputer is likely to remain an important resource in new research for years to come.
READ MORE: Energy & supercomputers: a data relationship by Michelle Leslie