Atos Launches First Supercomputer Equipped with NVIDIA A100 GPU

New Atos accelerator blade allows JUWELS BullSequana technique to turn out to be the most potent supercomputer in Europe

Paris, France, May well fourteen 2020

Atos, a international chief in electronic transformation, nowadays announces its new BullSequana X2415, the very first supercomputer in Europe to integrate NVIDIA’s Ampere subsequent-era graphics processing unit architecture, the NVIDIA® A100 Tensor Main GPU. This new supercomputer blade will produce unprecedented computing ability to increase software performance for HPC and AI workloads, tackling the challenges of the exascale era. The BullSequana X2415 blade will increase computing ability by far more than 2* and optimize electrical power intake many thanks to Atos’ a hundred% very productive water-cooled patented DLC (Direct Liquid Cooling) option, which takes advantage of warm water to awesome the device.

Forschungszentrum Julich Ralf-Uwe Limbach

Forschungszentrum Jülich will integrate this new blade into its booster module, extending its present JUWELS BullSequana supercomputer, earning it the very first technique throughout the world the use this new engineering. The JUWELS Booster will provide researchers throughout Europe with significantly increased computational resources. Some of the assignments it will fuel are the European Commission’s  Human Brain Project  and the Jülich Laboratories of  “Climate Science” and “Molecular Systems”. When thoroughly deployed this summer time the upgraded supercomputing technique, operated underneath ParTec’s program ParaStation Modulo, is expected to provide a computational peak performance of far more than 70 Petaflops/s earning it the most potent supercomputer in Europe and a showcase for European exascale architecture.

“By integrating the most effective systems offered, including this most current blade from Atos which contains NVIDIA’s potent A100 GPU into our supercomputer, we are in a position to increase our supercomputing ability. This is an critical milestone on our route to access our ultimate goal of furnishing exascale abilities to science and business” reported Prof. Thomas Lippert, Director of the Jülich Supercomputing Centre (JSC).

“We’re happy to be 1 of the very first supercomputer makers throughout the world to incorporate NVIDIA’s A100 GPU and to be in a position to provide researchers throughout the world the optimum software performance for their HPC and AI workloads reported Agnès Boudot, Senior Vice President, Head of HPC & Quantum at Atos.  “We’re actually fired up to be functioning with Jülich to increase its computing ability and to permit its JUWELS technique to turn out to be the most potent supercomputer in Europe.”

“Atos is 1 of the top rated gamers world-vast in the substantial conclude of the server marketplace. It is the major European-dependent vendor of HPC units, mostly thanks to the performance and ability of its BullSequana units, its potent technological skills and innovation and its potential to correctly handle huge-scale assignments globally. It has doubled its existence in the TOP500 ranking around the last 5 several years and its server marketplace share throughout the world carries on to expand. We see Atos as 1 of the key supercomputing makers that will possible advantage around the subsequent couple several years as desire for exascale units gets to be far more important.” said Steve Conway, Senior Advisor at Hyperion Research.

The new blade capabilities an NVIDIA HGX-A100 base-board outfitted with 4 NVIDIA A100 GPUs connected by using third-era NVIDIA NVLink™ engineering, along with two AMD EPYC CPUs and up to 4 NVIDIA Mellanox InfiniBand ports connected by using a Dragonfly+ configuration. The NVIDIA Ampere architecture is a groundbreaking engineering achievement that delivers a huge performance advancement around the former era for AI education and inference. The NVIDIA A100 Tensor Main GPU capabilities the world’s largest seven-nanometer processor, with far more than fifty four billion transistors. This crew of systems, alongside with the NVIDIA marketplace-major ecosystem of around seven-hundred GPU-accelerated HPC programs and assistance from NGC, will let researchers to hit the floor operating with containerized program optimized for the advancement and deployment of GPU-accelerated HPC and AI assignments.

The BullSequana X2415 will be offered in Q2 of 2020.



* in contrast to Atos’ present GPU blade


X2415 requirements

 1U variety variable blade with one acceleration node.

Direct Liquid Cooling.

Per node:

one x CPU Board with

  • 2 x AMD EPYC/Milan CPUs
  • 16 x DDR4 32GB @3200MTs minimum amount

one x GPU board “HGX-A100” with 4 x NVIDIA A100 GPUs

2 x Interconnect Mezzanine boards (BXI or HDR200)

Optional one x M.2 SATA SSD up to one.92TB or one x M.2 NVMe SSD 960GB


About Atos

Atos is a international chief in electronic transformation with 110,000 staff in seventy three international locations and annual earnings of € twelve billion. European range 1 in Cloud, Cybersecurity and Substantial-Performance Computing, the Team presents conclude-to-conclude Orchestrated Hybrid Cloud, Massive Information, Company Purposes and Digital Workplace methods. The Team is the Around the globe Information Technological know-how Associate for the Olympic & Paralympic Online games and operates underneath the models Atos, Atos|Syntel, and Unify. Atos is a SE (Societas Europaea), mentioned on the CAC40 Paris stock index.

The intent of Atos is to assistance design and style the upcoming of the information and facts place. Its skills and services assistance the advancement of knowledge, training and research in a multicultural strategy and lead to the advancement of scientific and technological excellence. Across the world, the Team allows its consumers and staff, and users of societies at huge to stay, function and produce sustainably, in a risk-free and secure information and facts place.

Push get hold of:

Laura Fau | [email protected] | +33 6 seventy three 64 04 eighteen | @laurajanefau