Supercomputing technologies help universities gain a competitive edge in their technical, engineering, and scientific applications. Academic use cases range from simulation tasks to data-intensive research, a diversity of workloads, which require universities to expand their processing capabilities or miss out on the next wave of discoveries. 

The Indiana University (IU) is a well-known institute that relies on the latest technologies and tools to facilitate their research. IU utilizes highly dense and parallel computing to expedite their activities and open up new performance capabilities that were once considered impossible. To achieve this, the university set out to build an AI-enabled system that would give researchers the power to explore new frontiers in medicine, cybersecurity, climate change, and more.

IU partnered with HPE to deliver Big Red 200, a high-density, water-cooled cluster system that contains nearly 100,000 cores and leverages the HPE Slingshot technology to deliver unmatched performance and energy efficiency. The system is the first of the HPE Cray EX supercomputers to be deployed and is specifically engineered for IU’s state-of-the-art data center. Big Red 200 features 672 compute nodes, each equipped with 256 gigabytes of memory and two 64-core AMD EPYC™ processors. IU is making continuous upgrades to the data center, which include warm water cooling and the addition of 64 AMD GPU-accelerated nodes, each with 512 gigabytes of memory. After this expansion, Big Red 200 is expected to operate at a peak rate of 6 petaflops, more than 6x faster than its predecessor.

Moving forward, IU has continued plans to expand their supercomputing capabilities with HPE to support more scientists, meet their research objectives, and foster economic development.

Supercomputing technologies from HPE are redefining how organizations approach weather and climate studies. HPE Cray systems are being used to power some of the most critical projects across the industry, helping scientists and researchers achieve better fidelity and faster predictability by accelerating key workloads.

CSC, the Finnish IT Center for Science, undertakes massive research projects across numerous scientific disciplines. Climate change is a top priority as rising sea levels and shifting weather patterns raise crucial questions about the safety and sustainability of our world. To take on these initiatives, scientists require top-notch computing resources to execute data-heavy and compute-intensive studies. 

CSC enlisted the help of HPE to build a leading-edge system to fast-track mission-critical simulations. LUMI, based on the HPE Cray EX, is one of the world’s fastest supercomputers with a peak performance of 552 petaflops per second. With its vast computational capacity, LUMI will execute workloads faster than previous systems, combining AI, deep learning, and sophisticated simulations with data analytics to fuel weather and climate research. The number-crunching capability of LUMI is accelerated by a partition of AMD Instinct GPUs and a planned future partition of 64-core AMD EPYC™ CPUs, making it one of the most advanced platforms for AI.

The U.S. Department of Energy’s Oak Ridge National Laboratory (ORNL) is also home to an HPE Cray EX supercomputing deployment. Comprising two HPE Cray EX supercomputers, the system delivers a combined 7.2 petaflops of numerical weather prediction capability to support the U.S. Air Force and Army missions worldwide. The system features 800 nodes, each equipped with two AMD EPYC™ processors, with future plans to enable heterogeneous compute. Each four-cabinet system can be expanded to 1,024 nodes, accommodating AMD Instinct GPU-based nodes that would boost throughput by 10x.

With extraordinary performance, the ORNL is breaking new ground in weather forecasting for a number of targeted scenarios—including forecasting stream flow, flooding, land submersion, and water depths. The end-to-end supercomputing technologies made possible by the HPE Cray EX can enable greater speed and dedicated performance to advance simulations in weather forecasting that were not possible before. The ORNL plans to accomplish this by implementing comprehensive cloud physics and creating a global hydrology model that involves simulating hundreds of watershed and drainage basins to increase the accuracy in predicting future events.

The manufacturing sector has grown increasingly more competitive as organizations work to accelerate the development of products and services. Organizations are applying computer-aided engineering (CAE), computer-aided design (CAD), and simulation techniques to hone and diversify their offerings to remain competitive, while revealing operational efficiencies to increase their bottom lines. Powering high-fidelity models is a top priority for manufacturers to gain real-time insight. The benefits of CAE/CAD technology are limitless, but only with the right infrastructure in place. The ideal production environment is flexible and cost-effective, enabling organizations to keep pace with industry trends and customer expectations. To accomplish these goals, manufacturers are exploring the potential of supercomputing to advance modeling and simulation. 

The use of HPC technologies like supercomputing is responsible for significant progress in the manufacturing sector, driving improvements across product design, testing, and development and meeting the rising demand for sustainable practices. HPE compute is extending the use of CAE/CAD capabilities, so today’s manufacturers can produce high-quality products and services quickly and accurately. 

HPE and AMD have teamed up with Ansys, a pioneer in simulation software, to help manufacturers uncover engineering insight with the power of supercomputing. These solutions enable organizations to implement technologies that can support the extreme scaling of demanding CAE applications to facilitate high-resolution models and more simulations. These capabilities provide immediate and predictive insight into how products like autonomous vehicles, aircrafts, and medical devices behave and perform under real-world conditions. Organizations that use HPE supercomputing technologies have the competitive edge to deliver safer, more effective products at low TCO.

Advancements in supercomputing are accelerating energy and gas exploration, allowing organizations to execute compute-intensive workloads at high speeds. These insights are crucial to pinpoint the location of energy reservoirs that exist underground or beneath the seafloor, as well as determine the ideal materials and methods to extract these resources for transportation and refinement. Traditional compute environments lack the performance to rapidly capture data and convert it into actionable intelligence, which can result in unplanned downtime, catastrophic failure, and lost revenue. Energy and gas organizations need to modernize their IT to improve their operations. 

A leading producer of the energy and chemicals has built its legacy by investing in cutting-edge supercomputing technologies to enhance their upstream capabilities and support a global downstream network. The business has made supercomputing innovation a top priority to solve issues related to exploration and discovery, such as worker safety and resource optimization, all the way to production. Their hope was to develop a new system that would allow them to dramatically reduce risks and promote efficiency with the use of sophisticated imaging and deep learning algorithms.

To reach this goal, the business unveiled the latest and most powerful addition to its supercomputing environment. The new system presents new opportunities in both exploration and development, with the ability to make more informed decisions on operations in the field and investments. The supercomputing system was created by HPE to deliver 55.4 petaflops of peak computing power, making it a top ten-ranked supercomputer on the TOP500®. This groundbreaking capability can process and image some of world’s largest geophysical datasets to generate detailed 3D earth models. As a result, the business can run simulations with dedicated seismic analytics designed to assist in the discovery of oil reservoirs, while reducing overall risks and costs by identifying when and where to drill.

Financial services institutions (FSI) make millisecond-decisions that affect every area of operations—from identifying and resolving cybersecurity threats to developing better financial products and accelerating intelligence to boost profit margins. 

Utilizing data from sources such as transactions, payments, and operational workflows, financial institutions can predict emergent trends, determine favorable stock trades, improve risk management, and enable fraud detection for real-time events. However, organizations operating on outdated technologies often struggle to efficiently capture and derive intelligence from the available data. To compete in the fast-paced FSI market, decisions must be made instantly and with precision. This presents a tremendous hurdle for organizations operating on legacy infrastructure that struggle to execute data-heavy workloads at scale.

Building resiliency, adapting to changing regulatory requirements, and developing more effective ways to engage customers will be key competencies moving forward. With the use of supercomputing, financial institutions can overcome the congestion and latency issues caused by traditional compute. HPE has already helped organizations evolve and reach new heights of financial performance with solutions that are designed to streamline core operations and fuel compute across any environment. We have delivered cutting-edge technologies and tools for FSI for over 35 years. And the HPE supercomputing technologies powered by AMD processors are the next stage of this important legacy, making supercomputing more accessible and affordable for organizations of all sizes. 

Data growth is impacting healthcare and life sciences in profound ways. Many organizations in these sectors are unable to manage the flood of data being generated by an increasing number of sources. The problem is compounded for those that lack the capacity to channel this information into deep insights. Compute technologies that have been used for the past decade are no longer powerful enough to analyze this crucial data, causing scientists, researchers, and caregivers to fall behind while organizations struggle to work and innovate.

To enter the next phase of medical and scientific advancement, healthcare and life sciences are transitioning to a new type of infrastructure to accelerate vital workloads and scale as data volumes continue to rise. Supercomputing is aiding discovery across these industries and revolutionizing what we know about ourselves and the world around us. Adopting an HPE Cray supercomputing system at the right scale with the right mix of AMD CPUs and GPUs allows organizations to select the right amount of compute for their specific use cases. Three primary applications include personalized genomics, biomedical imaging, and computational chemistry.

Researchers use supercomputing to visualize complex phenomena in the human body and process highly complex, granular, and unstructured data to understand biological and chemical interactions that can lead to new drug therapies and curing diseases. These technologies provide the essential power and flexibility required for high-speed analytics.

Computational chemists and researchers develop sophisticated programs to model and study the structures, functions, and interactions of biologically important molecules by relying on supercomputing performance to apply AI and machine learning techniques at scale. Supercomputers can derive insights from vast amounts of data to answer pressing questions in biochemistry.