NVIDIA announced that the JUPITER supercomputer, powered by the NVIDIA Grace Hopper
platform, is the fastest in Europe - delivering a more than 2x speedup for
high-performance computing and AI workloads compared with the next-fastest
system.
Soon capable
of running 1 quintillion FP64 operations per second, JUPITER is on track to be
Europe's first exascale supercomputer. The system enables faster simulation,
training and inference of the largest AI models - including for climate
modeling, quantum research, structural biology, computational engineering and
astrophysics - empowering European enterprises and nations to drive scientific
discovery and innovation.
Among the top
five systems on the TOP500 list of the world's fastest supercomputers, JUPITER
is the most energy efficient, at 60 gigaflops per watt.
Comprising
nearly 24,000 NVIDIA GH200 Grace Hopper Superchips and interconnected
with the NVIDIA Quantum-2 InfiniBand networking platform, JUPITER is
expected to reach over 90 exaflops of AI performance and is based on Eviden's
BullSequana XH3000 liquid-cooled architecture.
JUPITER also
incorporates NVIDIA's full stack of software for optimized performance.
"AI will
supercharge scientific discovery and industrial innovation," said Jensen
Huang, founder and CEO of NVIDIA. "In partnership with Jülich and Eviden, we're
building Europe's most advanced AI supercomputer to enable the leading
researchers, industries and institutions to expand human knowledge, accelerate
breakthroughs and drive national advancement."
Built
for Scientific Breakthroughs
Hosted by the Jülich Supercomputing Centre at the Forschungszentrum Jülich
facility in Germany, JUPITER is owned by the EuroHPC Joint Undertaking.
"With
JUPITER's extreme performance, Europe has taken a giant leap into the future of
science, technology and sovereignty," said Anders Jensen, executive director of
the EuroHPC Joint Undertaking. "JUPITER's computing power will serve as a
catalyst for scientific discovery, propelling foundational research across the
continent in fields as diverse as climate modeling, energy systems and
biomedical innovation."
"JUPITER is a
landmark achievement for European science and technology," said Thomas Lippert,
codirector of the Jülich Supercomputing Centre. "Powered by NVIDIA's
accelerated computing and AI platforms, JUPITER is advancing the frontier of
foundation model training and high-performance simulation, enabling researchers
across Europe to tackle challenges of unprecedented complexity."
"JUPITER will
substantially advance quantum algorithm and hardware development," added
Kristel Michielsen, codirector of the Jülich Supercomputing Centre. "Hybrid
quantum HPC-computation will profit from powerful tools such as the NVIDIA
CUDA-Q platform and the NVIDIA cuQuantum software development kit."
"JUPITER's
launch is not just an extraordinary technical success - delivering an exascale
machine and Julich's modular data center in less than nine months - it marks a
pivotal moment for European high-performance computing," said Emmanuel Le Roux,
senior vice president and global head of advanced computing at Eviden, Atos
Group. "It clearly demonstrated the technological leadership of the European
Eviden-led consortium, which designed, built and delivered this world-class
system."
Early testing
of JUPITER was conducted with the Linpack benchmark, which was also used to
determine the TOP500 ranking.
The JUPITER
supercomputer represents a new generation of computing systems, uniting
NVIDIA's end-to-end software stack to solve challenges in areas including:
- Climate and weather modeling: Enables high-resolution, real-time environmental
simulations and visualization, using the NVIDIA Earth-2 open platform. This contributes to
global community initiatives like the Earth Virtualization Engines
project, which aims to create a digital twin of the Earth to better
understand and address climate change.
- Quantum computing research: Advances quantum algorithm and hardware development
with powerful tools such as the NVIDIA CUDA-Q
platform and the NVIDIA cuQuantum software development kit.
- Computer-aided engineering: Reinvents product design and manufacturing through
AI-driven simulation and digital twin technologies, powered by the NVIDIA
PhysicsNeMo framework, NVIDIA
CUDA-X libraries and the NVIDIA
Omniverse platform.
- Drug discovery:
Streamlines the creation and deployment of AI models vital to
pharmaceutical research through the NVIDIA
BioNeMo platform, accelerating time to insight in biomolecular
science and drug discovery.