Workshop Overview
How can AI workloads be engineered for optimal performance in modern HPC environments?
The rapid advancement of Artificial Intelligence (AI) and Machine Learning (ML) has positioned High-Performance Computing (HPC) systems as indispensable platforms for developing, training, and executing these workloads. However, the architectural complexity and batch-oriented design of traditional HPC systems pose unique challenges distinct from those encountered in resource-elastic environments such as clouds. As AI models, such as large language, vision foundation, and multimodal architecture continue to grow in scale, their performance and scalability increasingly depend on HPC-grade architectures, advanced networking, and software engineering practices. The emergence of measurement-driven performance engineering and hybrid AI/HPC workflows is reshaping how large-scale computation is designed, benchmarked, and optimized.
The parallelization characteristics, input/output requirements, and dynamic workflows of AI workloads demand innovative techniques for efficient utilization of HPC resources. Moreover, the performance engineering of such workloads is crucial to achieve scalability, portability, and reproducibility across diverse system architectures.
Call for Papers
We invite submissions presenting experimental results, architectural insights, performance studies,
and best practices related to AI/ML workloads on HPC systems.
- Full papers: 4-10 pages (ACM style)
- Lightning talks: 1-page abstract (excluded from proceedings)
- Proceedings will follow SC guidelines
Topics of Interest
- Characterizing and optimizing AI/ML workloads on HPC
Characterizing AI/ML workloads on HPC systems, parallelization strategies for AI/ML, performance optimization of AI/ML frameworks on HPC, and measurement-driven performance engineering and hybrid AI/HPC workflows
- Integrating AI/ML into HPC environments
Best practices for integrating ML/AI into existing HPC environments, cross-platform portability and reproducibility in AI performance studies, and DevOps and MLOps for HPC-AI/ML
- Efficient execution of AI/ML on HPC
Efficient inference of LLMs on HPC, resource allocation and scheduling for AI/ML workloads, and energy efficiency and power management for AI/ML on HPC
- AI-enhanced HPC and scientific applications
AI-enhanced HPC simulations for scientific and industrial applications, HPC-AI/ML convergence for scientific applications, and industrial AI/ML on HPC
- Frameworks, benchmarking, and evaluation
Specialized AI/ML frameworks for HPC, HPC-AI/ML benchmarking and evaluation, and collaborative/interactive AI/ML on HPC
- HPC-AI infrastructure and architecture
Next-generation HPC systems for AI/ML, co-design of AI models and HPC architectures (accelerators, interconnects, memory hierarchies), and liquid cooling and power-constrained AI/HPC optimization
- Reliability, scalability, and development
Fault tolerance and resilience for long-running AI/ML training jobs, AI factories and end-to-end pipelines for scalable AI development on HPC, data preparation for AI/ML workload on HPC, and hybrid workloads on HPC systems
Important Dates
- Submission Deadline: June 15, 2026 (AOE)
- Acceptance Notification: August 31, 2026
- Camera-Ready Deadline: September 21, 2026 (AOE)
- Workshop Day: November 16/20, 2026 - to be announced
Organizers
- Dr. Siavash Ghiasvand – Dresden University of Technology, ScaDS.AI, Germany
- Dr. Vijeta Sharma – Norwegian University of Science and Technology, NTNU, Norway
- Dr. Ajeet Ram Pathak – Norwegian University of Science and Technology, NTNU, Norway