HPE signed the contract with EuroHPC for creation of the supercomputer worth $160 million

2020: Signing of the contract with EuroHPC on creation of the LUMI supercomputer for $160 million

On October 23, 2020 the company Hewlett Packard Enterprise (HPE) announced signing of the contract with European joint venture for development of high-performance computing (EuroHPC) on creation supercomputer for the amount more than 160 million dollars. The supercomputer under the name LUMI will be located in Finland and will help the European researchers, the private and public organizations to make break in scientifically- research works and to stimulate development of innovations in health care, meteorology and technologies based on artificial intelligence (AI).

According to the company, the LUMI supercomputer is a pre-eksaflopsny system with an expected peak capacity more than 550 PetaFLOPS (Pflops) that is equivalent to simultaneous operation of 1.5 million notebooks.

Within cooperation with EuroHPC the HPE company will expand resources of supercomputers for acceleration of implementation of the European plan of implementation of eksaflopsny calculations. It will become the following jump in the field of development of supercomputers.

We are very glad that HPE was selected for work on LUMI and our technologies of an eksaflopsny era will be involved in creation of one of supercomputers. We intend to continue the cooperation with EuroHPC directed to expansion of use of supercomputers of the next generation for support of scientific research, promotions of innovations and stimulation of economic growth. Peter Ungaro, the senior vice president and the general manager of division of high-performance computing and the crucial systems in HPE told

HPE will construct the pre-eksaflopsny supercomputer in To Europe the LUMI System will be equipped with supercomputers HPE Cray EX with processors AMD EPYC and AMD Instinct of the next generation that will provide the convergent workloads integrating modeling, simulation, AI and to the analyst Big Data, will accelerate carrying out difficult researches. The LUMI supercomputer will be placed in CSC – the Center IT for science – in the city of Kayaani in Finland. Its capacity are going to be provided to ten European countries within the created consortium LUMI. Enter into consortium: Belgium Denmark Iceland Norway Poland Finland Czech Republic Switzerland Sweden and Estonia. This system will become available in the middle of 2021.

HPE offers technologies of high-performance computing and AI within support of projects of EuroHPC on implementation of supercomputers of the next generation, including the pre-eksaflopsny and eksaflopsny systems. LUMI is the second project which partner selected HPE. Earlier the company announced participation in creation of the National supercomputer center IT4Innovations in the city of Ostrava which will become the supercomputer in the Czech Republic.

Moreover, HPE will expand the supply chain of supercomputers in Europe due to production of the HPE Cray EX supercomputers with liquid cooling and the HPE Apollo systems in the city of Kutna Hora in the Czech Republic. HPE will also create Competence Center in Europe which will have necessary tools and examination for development and applications testing, programs, processors and other solutions necessary for creation of programs of an eksaflopsny class.

Using these systems and EuroHPC technologies aims to support research and development (Research and Development) for development of science.

The LUMI supercomputer, within the EuroHPC project, will be created using the HPE Cray EX supercomputers which already found application in three future eksaflopsny systems of the U.S. Department of Energy: Aurora in Argonnsky national laboratory, Frontier in Ok-Ridzhsky national laboratory and El Capitan in Livermore National Laboratory. The HPE Cray EX supercomputers integrate in themselves technologies of calculations, storages, interconnect and the coolings created especially for an era of eksaflopsny calculations. All of them will be applied also in the LUMI supercomputer:

• High-performance computing with the built-in opportunities of AI uses EPYC AMD processors and the graphic processors AMD Instinct of the next generation;
• The expanded storage system for support of complex workloads of modeling, simulation and AI is created based on the HPE Cray ClusterStor E1000 system;
• Thanks to HPE Slingshot specially developed network for high-performance computing (interconnect) meets the requirements for speed and control of overloads (congestion management) of network during the loadings requiring transfer of a large number of data;
• Advanced opportunities of direct liquid cooling for the large-scale systems increase overall effectiveness of use of water and electricity at system operation.