The International Conference for High Performance Computing, Networking, Storage, and Analysis
The International Conference for High Performance Computing, Networking, Storage, and Analysis
Authors: Florian Berberich (Forschungszentrum Jülich, PRACE aisbl), John Towns (University of Illinois Urbana-Champaign, National Center for Supercomputing Applications (NCSA)), Toshiyuki Imamura (RIKEN Center for Computational Science (R-CCS)), Pete Beckman (Northwestern University)
Abstract: In recent years, the HPC ecosystem has undergone profound changes. In Europe, EuroHPC JU invested a lot to develop a world-class supercomputing ecosystem with a very strong focus on AI. The objective of this BoF is to give an overview of the current state of HPC activities from Europe, Japan, and the U.S., with a particular focus on investments in public AI infrastructure. We will present and discuss with the different international HPC stakeholders the current state of play, future plans, challenges, and analyze critically the impact of an AI-focused strategy for the HPC ecosystem in general.
After a general introduction, there will be short contributions from the international HPC organizations presenting the current development of the HPC and AI ecosystem in their respective regions. A discussion with the audience will show different perspectives from industry and research. We will address hardware, software challenges, and identify gaps. The BoF will be closed with a summary of the results for preparing recommendations.
Long Description: In recent years, the HPC ecosystem in Europe and around the world has undergone profound changes. In Europe, e.g. EuroHPC JU made significant investments to develop a World-Class Supercomputing Ecosystem with recently a very strong focus on AI (AI (Giga-)Factories). In the U.S., the National Artificial Intelligence Research Resource (NAIRR) connects U.S. researchers to responsible and trustworthy AI resources, as well as the needed computational, data, software, training, and educational resources to advance research, discovery, and innovation. The objective of this BoF is to give an overview of the current state of HPC activities from Europe, Japan and U.S., with a particular focus on investments in public AI infrastructure.
For a successful HPC ecosystem the requirements of the different HPC users (simulation, ML and AI) need to be satisfied. This is reflected in a diverse compute architecture being able to accommodate FP64 stimulation workloads as well as FP8 training workloads for the AI community.
A special focus of this BoF is the current strong AI orientation and it`s impact on the use of the AI oriented systems; additionally, AI`s strong impact on other domains.
The BoF will be divided in two parts. In the first part, several key HPC organisations from Europe, Japan and U.S. will present in short contributions their current state and plans for future HPC investments.
The second part will focus on discussions regarding the impact of AI on industrial and academic research. The different approaches and offers will be compared and paths for cooperation will be investigated. We will discuss and compare the EU/Japan/U.S. investments in AI.
The current development is twofold: new investments in HPC bring as well new opportunities, however only research fields having sound experimental or synthetic data available could benefit from AI methods. In addition, HPC vendors will follow the market and concentrate on AI optimized processors which will potentially reduce the available variety of processors.
The second part will end with an interactive round table discussion targeting the assessment of the integration of actors, of the investments in AI towards a well-integrated, resilient, world-class HPC ecosystem. The discussion will also aim at identifying projects to deploy that vision and structures to guarantee the delivery of a persistent set of services beneficial to the academic and industrial users of HPC. The BoF participants are encouraged to contribute to the discussion by ranking different services and identifying gaps and shortcomings. The result will be used in order to prepare recommendations for HPC roadmap exercises and to provide input to the national and international HPC ecosystem policy makers.