Advanced Computing in the Age of AI | Sunday, May 26, 2024

NVIDIA Powers AI Supercomputing, Scientific Innovations, and HPC Research 

NVIDIA, the world leader in accelerated computing,  is driving a fundamental shift in the high-performance computer (HPC) industry. This week NVIDIA detailed how its AI chips are accelerating innovation for AI-powered systems for advanced supercomputing and scientific discovery.  

The chipmaker announced at the ISC High Performance event in Hamburg that new supercomputers around the globe are using NVIDIA Grace Hopper Superchips (GH200). This includes supercomputer JUPITER at the Jülich Supercomputing Centre in Germany, CEA at the French Alternative Energies and Atomic Energy Commission, and Alps at the Swiss National Supercomputing Centre from Hewlett-Packard Enterprise (HPE). 

According to NVIDIA, the supercomputing systems using Grace Hopper chips represent a combined total of 200 exaflops of computational AI processing power. This is equivalent to a staggering 200 quintillion calculations per second.

“AI is accelerating research into climate change, speeding drug discovery, and leading to breakthroughs in dozens of other fields,” said Ian Buck, vice president of hyperscale and HPC at NVIDIA. “NVIDIA Grace Hopper-powered systems are becoming an essential part of HPC for their ability to transform industries while driving better energy efficiency.”

NVIDIA has also announced its goal to accelerate its quantum computing efforts, which include the installation of the open-source NVIDIA CUDA-Q platforms at several international sites. 

CUDA-Q is an open-source accelerated computing platform that powers quantum processing units (QPUs), specialized and designed to perform quantum computations that are too computationally intensive for classical computing. 

QPUs enable researchers to perform complex quantum simulations and optimization tasks. In addition, it allows them to leverage quantum parallelism for rapid computations and enhance machine learning algorithms to accelerate scientific discovery and innovation. 

“Useful quantum computing will be enabled by the tight integration of quantum with GPU supercomputing,” said Tim Costa, director of quantum and HPC at NVIDIA, in a statement. “NVIDIA’s quantum computing platform equips pioneers such as AIST, JSC, and PSNC to push the boundaries of scientific discovery and advance the state of the art in quantum-integrated supercomputing.”

In March this year, NVIDIA introduced the Earth-2 platform for predicting global climate using simulation by AI supercomputers. The platform, which is part of NVIDIA CUDA-X microservices software, has the potential to provide highly accurate insights into weather fluctuations and provide timely warnings to mitigate the rising threat of climate disasters around the globe. 

The power and efficiency of NVIDIA GPUs are also being utilized at Argonne National Laboratory, a multidisciplinary science and engineering research center in Illinois, to generate gene sequences that help them better understand the virus behind COVID-19 and accelerate medical discoveries.  

Earlier this year, Microsoft expanded its partnership with NVIDIA with powerful new integrations with NVIDIA technologies to advance AI infrastructure and accelerate research in healthcare and life sciences. 

The technological advancements enabled by NVIDIA are busting bottlenecks in AI supercomputing, HPC research, and scientific discoveries. From predicting global climate change patterns to the discovery of new materials, the contributions of NVIDIA technologies to the scientific community are profound. As NVIDIA continues to roll out more advanced technologies, we can expect the chipmaker to play a key role in the next wave of scientific breakthroughs and transformative innovations.

Related Items 

How AI Will Impact the Drug Discovery Pipeline 

Scientists Use Machine Learning To Prove The Existence of a Rare Phase Of Matter 

Stanford HAI AI Index Report: Science and Medicine