Artificial intelligence is rapidly emerging as a transformative technology that will affect science, engineering, industry, public administration, and many other sectors of society. At HLRS, we offer a powerful computing infrastructure and technical expertise that make it possible to explore the possibilities that artificial intelligence now offers. Our supercomputing capabilities support applications using generative AI, deep learning, large language models, and foundation models. Scientists in our Department of Converged Computing are also improving methods for integrating AI, simulation, cloud computing, and other advanced computing technologies in efficient workflows.
Beginning in early 2025, this next-generation, APU-accelerated supercomputer will offer increased performance for AI and hybrid computing.
HLRS's current flagship supercomputer supports large-scale data analytics as well as new kinds of hybrid workflows that combine HPC and AI.
This partition of HLRS's Vulcan cluster is ideal for smaller-scale AI applications; for example, for SME's that require only a single, mid-range GPU per node.
Integrated into our Vulcan cluster, this system is dedicated to research related to the COVID-19 pandemic and other needs for crisis computing.
HLRS provides computing infrastructure and expertise in AI for scientists, industrial researchers, and organizations of other kinds interested in testing new applications of AI technologies. Click for examples of how our systems are used.
In addition to providing access to high-performance computing systems, HLRS staff has expertise in supporting methods and concepts that are driving the AI field forward.
The future of high-performance computing lies in hybrid approaches that integrate AI with HPC, cloud, and edge computing technologies.
Incorporating physical constraints in machine learning methods can improve model accuracy and utility.
Our systems make it possible to train large language models like ChatGPT, and to use such models in generative AI applications.
We enable adaptation of models based on large, diverse datasets to develop customized solutions to specific problems.
Are you interested in improving your knowledge about programming artificial intelligence applications on high-performance computing systems? HLRS regularly offers courses in machine learning, deep learning, and related topics that can help you turn your ideas into AI solutions.
October 14 - 31, 2024
Online (flexible)
November 04 - December 13, 2024
March 10 - 12, 2025
Online
HLRS is a founding member of the Smart Data Solution Center Baden-Württemberg, a membership organization that brings together researchers, users, and companies offering data analytics solutions to pool expertise and accelerate the development of shared solutions for commonly faced problems. (Click to visit sdsc-bw.de.)
In addition to providing hardware and solutions for AI, research scientists at HLRS coordinate and contribute to collaborative, funded research projects that are addressing key technical problems facing the AI field and exploring potential applications of AI to solve pressing challenges. These activities enable us to build expertise and help us to better address the interests and needs of our system users.
The HiDALGO2 project is addressing challenges caused by climate change, focusing on technical issues related to scalability on HPC and AI infrastructures, the use of computational fluid dynamics methods, and uncertainty analysis.
HLRS is the coordinating center for this project to integrate artificial intelligence (AI) topics into curricula at the University of Stuttgart, and to implement AI technologies to improve instruction.
DECICE is developing an open and portable cloud management framework that will enable the automatic and adaptive optimization of software applications for heterogeneous computing architectures.
See a list of all HLRS projects related to artificial intelligence.
Would you like to learn more about using HLRS's systems for artificial intelligence? Please contact us.
Head, Converged Computing
Aug 06, 2024
Jul 05, 2024
Jun 21, 2024
May 24, 2024
