Supercomputing

Reach New Scientific Heights

Predicting weather. Discovering drugs. Finding new energy sources. With performance that can reach—and even exceed—petascale performance, supercomputing gives researchers the power they need to simulate and predict our world. Improving productivity and increasing the number of scientific simulations can have a profound impact on the quantity and quality of their scientific breakthroughs.

POWER FOR THE WORLD'S FASTEST SUPERCOMPUTERS

Today, NVIDIA GPUs provide the power for the fastest supercomputers in the US and Europe. In the US, Oak Ridge National Labs' Summit combines high-performance computing (HPC) with AI for over 200 petaFLOPS of double-precision computing for HPC and 3 exaFLOPS of mixed-precision computing to drive scientific discovery.

In addition, supercomputing data centres around the world are leading the NVIDIA Ampere architecture. They are using it to bring science into the exascale era and simulate larger models, train and deploy deeper networks, and advance an emerging hybrid field of AI-powered simulations.

Full power: HPC data centres accelerate
scientific work with A100 GPUs

SCIENCE IN ACTION

MAPPING THE INTERIOR OF THE EARTH

Princeton University's Department of Earth Science uses Summit with NVIDIA Volta GPUs to observe and simulate seismic data to image the Earth's interior around the world.

LIMITLESS ENERGY PRODUCTION

The Princeton Plasma Physics Laboratory is using Oak Ridge National Lab's Summit supercomputer with NVIDIA Volta GPUs to simulate and predict plasma properties for the next fusion reactor.

MAKING PERFORMANCE MORE EFFICIENT

Georgia Tech aims to implement maximum performance with minimum instability. Summit's computational power drives more reliable performance, more salient flow models for multi-scale and multi-physics simulation.

SOLVING GENETIC PUZZLES

From sustainable bioenergy to better understanding disease, Oak Ridge National Lab's Biosciences Division counts on Summit supercomputers with the full power of NVIDIA V100 Tensor Core GPUs to solve previously unsolvable fundamental biological problems.

RESEARCH SUPERNOVA EXPLOSIONS

Oak Ridge National Lab's Scientific Computing group is harnessing the incredible power and memory of NVIDIA Volta GPUs to explore supernova explosions and new parts of the nuclear landscape.

ENABLING SCIENTIFIC INNOVATION

From earthquakes and clean energy to supernovas, the DOE (Department of Energy) is enabling scientists at the Oak Ridge Leadership Computing Facility to bring their ideas to life on the GPU-powered Summit supercomputer.

RUNNING GPU-ACCELERATED APPLICATIONS

From weather forecasting and materials science to wind tunnel simulations and genomics, NVIDIA GPU accelerated computing is at the heart of the most promising research areas using HPC.

The NVIDIA CUDA® programming model is the platform of choice for developers of high-performance applications. More than 700 applications , including developers of the top 15 HPC applications, now have GPU support. Many of the best HPC applications are deployed as pre-configured containerised software on NGC.

View HPC application performance

Discover HPC software

Discover containers available in NGC

INNOVATION IN SCIENTIFIC FIELDS

WEATHER AND CLIMATE

Various weather and ocean modelling applications such as the Weather Research and Forecasting (WRF) model for weather research and forecasting and tsunami simulations achieve a huge speed-up of more than 5 times, saving time and improving accuracy.

COMPUTATIONAL CHEMISTRY

NVIDIA® Tesla® GPUs enable researchers to drive discovery in new dimensions. A single GPU node can replace multiple CPU cluster nodes. Compared to CPUs, GPUs are more than 5 times faster for common molecular dynamics, quantum chemistry, visualisation and docking applications. They also offer higher simulation performance per dollar and watt - for faster, deeper and more cost-effective insights.

PHYSICS

From fusion energy to high-energy particles: Physics simulations are applicable to a very wide range of applications in the HPC data centre. Compared to CPUs, GPUs accelerate the most important physics applications by more than 10 times and enable insights that were previously not possible.

SOLUTIONS TO ACCELERATE SCIENTIFIC DISCOVERIES

ACCELERATE ANY WORKLOAD

The NVIDIA A100 Tensor compute unit GPU delivers unprecedented acceleration at any scale for AI, data analytics and HPC to tackle the world's toughest computing challenges.

MORE INFOS

UNPRECEDENTED COMPUTING POWER

NVIDIA DGX-A100 is the universal system for all AI workloads, delivering unprecedented compute density, performance and flexibility in the world's first 5-petaFLOPS AI system.

MORE INFOS

CERTIFIED PERFORMANCE

NVIDIA Certified Systems™, which include both NVIDIA HGX™ and EGX™ server platforms, enable organisations to securely deploy scalable hardware and software solutions and optimally run their AI workloads.

MORE INFOS

PRACTICAL TRAINING

Hands-on training for AI, accelerated computing and accelerated data science Developers, data scientists, researchers and students gain practical experience using GPUs in the cloud. And IT professionals can access courses on building and managing infrastructure to support AI, Data Science and HPC workloads in their organisations.

Get started today with self-paced online training for individuals, instructor-led workshops for teams, and downloadable course materials for instructors.