Labs to Deliver Paper, Lecture and Workshop Discussion at Cray User Group Conference

[Labs_Editorial]

Every year, the Cray User Group (CUG) conference brings together hundreds of enthusiasts to share ideas, research and technological advancements in the world of High-Performance Computing (HPC).

This year’s event, running from May 4-8, will include an active presence from Hewlett Packard Labs. Researchers working at and with Labs will present a paper on quantum computing’s intersection with HPC, give a lecture on digital twins for data centers, and give a talk in a workshop about executing AI-HPC workflows in federated workflow environments.

The 2025 CUG Conference is being co-hosted by the HPE Cray User Group and the Institute for Advanced Computational Science (IACS) at Stony Brook University.

Dr. Robert J Harrison, one of the conference organizers, said the 2025 is exploring the theme of “Computing Horizons.”

“With this theme, we want to capture the sense of the great opportunities arrayed in front of us, with the tight integration of traditional HPC workloads and powerful AI enabled by innovations in algorithms and special purpose hardware,” said Harrison, IACS’s director at Stony Brook. “The 2025 conference is about this bright future and the exploration of these limitless possibilities and their associated challenges.”

Larry KaplanPavana PrakashXin ZhanXin Zhan, a Research Software Engineer with Hewlett Packard Labs, outlines the growing needs for scalable High Performance Computing (HPC) and Quantum Computing (QC) integration in her paper, “A Full Stack Framework for High Performance Quantum-Classical Computing.”

In the paper she presents her research lab’s HPC-QC full stack framework and its hybrid workload development capability with modular hardware/device-agnostic software integration approach. She explains how the lab’s work provides the framework for a unified quantum-classical programming environment built upon a classical HPC software stack (compilers, libraries, parallel runtime and process scheduling).

Pavana Prakash, a Research Scientist working on Applied AI at Labs, is presenting her group’s work addressing the challenges of using digital twins to help manage the integration of data centers and smart grids. She will outline how Labs is facilitating new ways to collect the right data so digital twins can support smart grids and GPU data centers in a more insightful and energy-efficient manner.

Directly after CUG, a related workshop will be held, “Expanding Horizons in AI and HPC.” This workshop includes a presentation by Larry Kaplan, a Senior Distinguished Technologist and Chief Software Architect for HPC who oversees the entire Cray supercomputer software stack. The presentation is on functionality being developed with Labs colleagues Aalap Tripathy and Martin Foltin on initiatives to make it easier to provision, monitor and reproduce HPC workflows in a federated environment.

In addition, members of HPE’s HPC/AI EMEA Research Lab contributed to a paper and a presentation at the conference.

The paper describes an innovative technique to drastically reduce the size of LLMs, resulting in better performances and energy efficiency. Contributors include HPE’s Alessandro Rigazzi along with German startup Seedbox, Advanced Micro Devices and the High-Performance Computing Center Stuttgart (HLRS).

The presentation focuses on using different MPI implementations on HPE Cray EX supercomputers for native and containerized applications execution. Contributors include HPE’s Alfio Lazzaro along with the University of Oslo and the Academic Computer Centre Cyfronet of the AGH University of Krakow.

Hewlett Packard Enterprise acquired Cray Inc., formerly Cray Research, in 2019.