Special Issue on Leveraging Cutting-Edge High Performance Computing for Large-Scale Applications Submission Date: 2025-09-30 Motivation and Scope

High Performance Computing (HPC) recently entered into the exascale era, marking an important milestone of its history. High-end supercomputers and clusters with remarkable level of performance are now commonly available for standard and specific computational needs, thereby increasing the focus on HPC and related topics. Leveraging the potential of large-scale supercomputers is an HPC skillful task that requires in-depth knowledge on both hardware and software. Indeed, the architectural structure of cutting-edge HPC processors is rather complex, with each feature provided by a specialized mechanism, the processing overhead of which can turn out to be an efficiency bottleneck. This challenge becomes more pronounced as the potential processing power of the machine increases. For example, at the compute node level, processors with many cores might have a NUMA architecture, a deeper memory hierarchy and/or wide SIMD capabilities. At the interprocessor level, hardware support for data exchanges can also significantly increase running time. There is a wide range of applications that genuinely need a whooping computing speed, one of them being the training of large-scale AI models. The evolution of HPC appears closely tied to the growing demand for speed from such large-scale applications. As a result, the implementation of cutting-edge techniques should remain scalable on large-scale machines. Achieving this level of efficiency is challenging but highly sought after by HPC community, including end-users.

This special issue aims at addressing the aforementioned technical context and scientific challenges, thus expecting to gather insightful contributions on how to leverage the increasing potential of HPC systems for cutting-edge applications. We invited various types of submissions: research papers (fundamental and experimental), position papers, extended version of conference papers, and surveys. Papers are solicited on a broad range of topics including (but not limited to):

Scalability with large-scale parallel machines (issues, limits, methods, applications)

Efficiency with large clusters of accelerators (e.g. FPGAs, GPUs, AI accelerators)

Parallel algorithms and scheduling for heterogeneous and/or hierarchical systems

Fault-tolerance for large-scale HPC systems

Models and tools for performance/energy evaluation on large-scale HPC systems

Power consumption and carbon footprint of large-scale HPC solutions

Applications that need cutting-edge HPC

Combinatorial optimization from the exascale standpoint

AI in the exascale era

Large-scale Cloud systems

Quantum computing (efficiency perspectives)

Integrating Classical and Quantum Computing

Guest Editors

Claude

Claude TADONKI

Mines Paris, PSL University, Paris, France

claude.tadonki@minesparis.psl.eu

mencagli

Gabriele MENCAGLI

University of Pisa, Pisa, Italy

gabriele.mencagli@unipi.it

Gabriele

Leonel SOUSA

Instituto Superior T´ecnico, Universidade de Lisboa, Lisboa, Portugal

las@inesc-id.pt

Important Dates

Submission portal opens: April 14, 2025

Deadline for paper submission: September 30, 2025

Latest acceptance deadline for all papers: December 30, 2025