6th Energy Efficient High Performance Computing State of the Practice Workshop
In conjunction with SCA/HPCAsia 2026
📅 January 26, 2026
🕤 9:30 – 16:30 (JST)
📍 Osaka, Japan 

The Energy Efficient HPC State of the Practice workshop addresses the operational, infrastructural, and environmental challenges involved in deploying and managing modern high-performance computing systems. Its primary objective is to collect and disseminate best practices, case studies, and reproducible operational experiences from HPC centers, facilities, and vendors worldwide.

Abstract

This presentation examines CINES’ approach to deploying and optimizing its supercomputing infrastructure, with a focus on the Adastra system and its sustainability challenges.

The session begins with a detailed look at data center design and the 2021 infrastructure upgrade, including cooling choices, compliance with French heat reuse regulations, and PUE and WUE evaluations.

It also explores new opportunities for integrating technologies to improve efficiency and continuously enhance the infrastructure with sustainability in mind.

The talk then addresses the call for tenders process, highlighting key lessons from past experiences and how energy and CO₂e requirements influenced vendor selection and system architecture. The discussion emphasizes total cost of ownership optimization, benchmark definition strategies, and the importance of refining architectural needs.

Finally, the presentation shifts to operational perspectives, covering machine tuning, the impact of frequency capping at scale, and cooling strategies. It also explores energy-aware and climate-aware scheduling, demonstrating how these practices reduce environmental impact and can significantly limit cooling infrastructure demands.

By sharing empirical data and best practices, this session provides actionable insights for HPC centers striving to balance performance, cost efficiency, and sustainability.

Bio

Gabriel Hautreux is in charge of the HPC and AI department at CINES, one of the three French national centers, and managed the deployement of the leading edge cluster Adastra, #11 at TOP500 and #3 at GREEN500 in November 22.

He previously worked as an HPC application expert, in charge of the French High Level Support Team (HLST) and enabled dozens of scientific teams to prepare their applications for upcoming systems in GENCI’s “Technological Watch Group” between 2016 and 2019.

His current activities aim to enable the research community to leverage application using exascale technologies, as well as increasing the energy efficiency of applications and reducing the global energy and carbon footprint of HPC center. 

He is a specialist in exascale architectures, new development paradigms and energy efficiency in HPC and AI systems.