Special Issue on High-performance Computing Heterogeneous Systems and Subsystems Submission Date: 2025-05-30 Motivation and Scope

High-performance computing (HPC) is at a pivotal juncture, characterized by significant advancements in computing technologies and architectural features. This special issue explores this dynamic field's latest advancements, challenges, and innovations. Heterogeneous HPC systems integrate diverse computational resources, including CPUs, GPUs, FPGAs, and other specialized accelerators, to deliver superior performance for various applications. To harness their potential fully, these systems require novel resource management, scheduling, programming models, and performance optimization approaches. Combining cutting-edge research and practical insights, this special issue provides a comprehensive overview of heterogeneous HPC systems' current state and future directions. It is a valuable resource for researchers, practitioners, and policymakers interested in leveraging heterogeneous computing to solve complex scientific, engineering, and data-intensive problems more efficiently and effectively.

The topics include but are not limited to:

1. Heterogeneous Programming Models and Runtime Systems:

Models, parallel resource management, and automated parallelization

Algorithms, libraries, and frameworks for heterogeneous systems

2. Heterogeneous Architectures:

Power/energy management, reliability, and non-von Neuman architectures

Memory and interconnection designs

Data allocation, caching, and disaggregated memory

Consistency models, persistency, and failure-atomicity

3. Heterogeneous Resource Management:

System and software designs for dynamic resources

High-level programming, run-time techniques, and resource frameworks

Scheduling algorithms, resource management, and I/O provisioning

4. Heterogeneity in Artificial Intelligence:

AI/ML/DL predictive models and optimized systems for heterogeneous workflows and applications

Tools and workflows for AI/ML/DL in scientific applications

Guest Editors

Sergio Iserte

Barcelona Supercomputing Center, Spain

sergio.iserte@bsc.es

Pedro Valero-Lara

Oak Ridge National Laboratory, USA

valerolarap@ornl.gov

Kevin A. Brown

Argonne National Laboratory, USA

kabrown@anl.gov

Important Dates

Submission portal opens: October 01, 2024

Deadline for paper submission: May 30, 2025

Latest acceptance deadline for all papers: July 31, 2025