The Center for High Performance Computing at the University of Stuttgart (HLRS) has launched the supercomputadora Hunter. Designed and manufactured by Hewlett Packard Enterprise (HPE), offers a world class infrastructure for large-scale simulation, artificial intelligence and data analysis applications for the scientific, industrial and public administration sectors.
“Hunter offers scientists at the University of Stuttgart and throughout Germany a future-ready infrastructure for AI-based simulations and high-performance computing of a new quality”explained Professor Peter Middendorf, rector of the University of Stuttgart, in the announcement. “Hunter also benefits our entire university ecosystem with its global players, strong mid-sized companies, and growing startup scene.”.
The €15 million supercomputer is funded by the German ministries of Science, Education and Research and will support a variety of research efforts. public and private research ranging from engineering and meteorological modeling to biomedical research, materials science. And of course everything related to AI in full expansion: “The rapid development of AI and the increasing focus on sustainability in supercomputing mean that high-performance computing is currently undergoing an exciting and transformative period.”they explain.
The Hunter supercomputer is underway
The Hunter supercomputer is based on the same HPE Cray architecture Supercomputing EX4000 which is used in the world’s three verified exascale systems. Each of its 188 nodes is equipped with four high-performance HPE Slingshot interconnects. The machine leverages HPE Cray Supercomputing E2000 storage systems, which are specifically designed to meet the demanding input/output (I/O) requirements of large-scale supercomputers, and HPE Cray Supercomputing Programming Environment software, which offers a suite full of tools to develop, port, debug and tune applications at scale.
HPE Performance Cluster Manager is also used for system health monitoring and management, as well as power management. Additionally, the entire Hunter system architecture uses 100% direct liquid cooling fanless, a first-of-its-kind technology developed by HPE and designed to improve the energy and cost efficiency of large-scale AI deployments.
With a theoretical maximum performance of 48,1 petaflops (48.1 billion floating point operations), Hunter’s speed is almost twice that of the German HLRS’s previous flagship supercomputer, called Hawk. At the same time, Hunter represents a major change in computer architecture. While previous systems relied primarily on a large number of CPUs to achieve high performance, Hunter relies on the AMD Instinct MI300A Accelerated Processing Units (APUs), which combine CPU, GPU accelerators, and high bandwidth memory. band in one package. Despite the higher performance, this architecture consumes 80% less energy.
The Hunter supercomputer is important for Germany and Europe, and is a transition designed to help researchers prepare and optimize their code for APUs and GPUs in anticipation of the larger Herder supercomputer, which will reach the top. -10 among the largest supercomputers in the world.
