Danish Centre for AI Innovation taps Eviden and NVIDIA for AI supercomputer
The Danish Centre for AI Innovation has selected Eviden and NVIDIA to help the institute build a new AI supercomputer, to support its ambitions in advancing the use of AI in the nation’s business landscape.
Denmark’s recently established national Centre for AI Innovation aims to provide researchers from public and private sectors access to an advanced supercomputer that can handle the latest and greatest in AI developments. The aim is to expedite research and innovation across various sectors such as healthcare, life sciences, and the green transition, fostering the creation of pioneering solutions to address pressing challenges.
The centre has been concepted and backed by the Novo Nordisk Foundation and the Export and Investment Fund of Denmark, and comes in response to the fact that Denmark currently does not have a GPU-accelerated supercomputer in operation.
“This partnership is therefore very important for Denmark’s research driven ecosystem,” said Mads Krogsgaard Thomsen, CEO of the Novo Nordisk Foundation. “It will help Denmark’s brilliant researchers and innovators rise to the next level.”
Eviden, a business of Atos, has been selected to design and develop the new supercomputer. The processors will be provided by NVIDIA.
Commenting on the project, Cédric Bourrasset, Head of HPC-AI and Quantum Computing at Eviden, said: “We’re proud to be working with the Danish Centre for AI Innovation to deliver the AI supercomputer to provide an unprecedented potential to accelerate groundbreaking advancement across sectors.”
“This new supercomputer will be a powerful resource to further boost AI research and collaboration between academia and Danish industry in Denmark’s already thriving AI-based research ecosystem.”
Notably, Eviden – which builds such supercomputers all over the world – said the Danish version will be among the most powerful AI supercomputers on the planet. Based on the NVIDIA DGX SuperPODTM architecture, it will deliver 6 Exaflops of FP8 AI performance.