Energy and Resource Efficiency of Data Centers

Energy Management in HPC Data Centers

ROOM 3

Tuesday September 19, 2023

Time: 14:00 - 18:30 h.

Schedule

Chair(s):
  • Maximilian Höb, Leibniz Supercomputing Centre, Munich, Germany
  • Dieter Kranzlmüller, Leibniz Supercomputing Centre, Munich, Germany
  • Andrew Grimshaw, University of Virginia, Charlottesville, USA
Information

Data centers have become an essential part of modern life, powering complex applications and enabling seamless connectivity. From internet searches and weather forecasts to online ordering, data centers play a crucial role. Scientific research, including data-intensive high-performance computing (HPC) and cloud computing, relies heavily on these facilities. The environmental impact of data center operations cannot be ignored, particularly in the context of energy consumption.

In many regions, the demand for computing and storage capacities surpasses available resources, necessitating an ecologically, economically, socially, and technologically conscious approach to development. Today, energy-efficient systems offer a viable solution to address this challenge, enabling the realization of more powerful and environmentally friendly data center infrastructures. By focusing on metrics such as performance per computing operation (Watts/FLOPS), energy efficiency can be significantly improved.


Efforts to optimize energy management in HPC data centers go beyond primary cooling solutions. Waste heat reuse and energy optimization are critical areas of focus. Real-time processing of sensor data using machine learning algorithms plays an increasingly significant role. Moreover, the regular replacement of hardware with more energy-efficient and powerful systems contributes to environmental impact through production and disposal.


Managing data centers efficiently requires complex optimization strategies to enhance resource usage and increase economic and environmental sustainability. Software plays a crucial role in these optimizations, as it controls energy-saving modes, data transfer, and computing operations. Container and workflow frameworks already enable more flexible and efficient hardware utilization, although their integration into supercomputing clusters is still in progress.

This workshop will explore various topics related to energy management in HPC data centers, including but not limited to:

  • Green Data Center IT and Software
  • Optimizing the energy efficiency of systems
  • Models for energy and resource efficiency assessment
  • Resource conservation through smart technologies and artificial intelligence
  • Real-time energy consumption prediction and regulation using machine learning
  • Efficient facility infrastructures and hybrid cooling technologies
  • Secondary use of waste heat and integrated community heating systems
  • Scalable, hybrid, and flexible compute and data system architectures
  • Container and workflow frameworks enabling energy efficiency methods
  • Green coding approaches in high-performance software
  • Scheduling algorithms for optimal energy use

Contributions that extend beyond these topics and align with the core theme of the workshop are particularly welcome. Join us to explore innovative approaches for achieving energy efficiency and sustainability in HPC data centers, paving the way for a greener and more efficient future.

 

Workshop Paper Submission

Contributors should consult Springer's authors' guidelines and use their proceedings templates, either for LaTeX or for MSWord, for the preparation of their manuscripts.

  • All submissions must be in English.
  • Regular articles should not exceed 15 pages (including references).
  • Articles must be formatted in the LNCS style of Springer Verlag LNCS style (see details at: https://www.springer.com/gp/computer-science/lncs/conference-proceedings-guidelines)
  • All submission reviews are Single-blind.
  • All submitted research articles will be peer-reviewed by at least three expert reviewers.
  • Only contributions that have not been submitted elsewhere or currently under review will be considered.
  • Only accepted contributions presented at the conference will be included in the proceedings.
  • Article submission is handled electronically with Springer’s Online Conference System: OCS (https://ocs.springer.com/ocs/home/CARLA2023)

 

Important Dates

Submission deadline     : July 30, 2023
Author notification         :  August 18, 2023
Workshop date              : September 19, 2023
Camera Ready papers  : October 6, 2023

All articles accepted for publication will be included in the CARLA 2023 Proceedings, to be published in Springer CCIS Series — Communications in Computer and Information Science.

 

Program Committee

  • Adam Belloum, University of Amsterdam, Amsterdam, The Netherlands
  • Marian Bubak, Sano Centre for Computational Medicine and ACC Cyfronet AGH, Krakow, Poland
  • Harold Castro, Universidad de los Andes, Bogotá, Colombia
  • Ewa Deelman, University of Southern California, Marina Del Rey, United States of America
  • Andrew Grimshaw, University of Virginia, Charlottesville, USA
  • Ladislav Hluchy, Slovak Academy of Sciences, Bratislava, Slovakia
  • Maximilian Höb, Leibniz Supercomputing Centre, Munich, Germany
  • Marek Kasztelnik, ACC Cyfronet AGH, Krakow, Poland
  • Dieter Kranzlmüller, Leibniz Supercomputing Centre, Munich, Germany
  • Cerlane Leong, Swiss National Supercomputing Centre CSCS, Lugano, Switzerland
  • Jan Meizner, Sano Centre for Computational Medicine and ACC Cyfronet AGH, Krakow, Poland
  • Martin Schulz, Technical University of Munich, Munich, Germany
  • Dan Stanzione, Texas Advanced Computing Center, Austin, United States of America

 

FURTHER INFORMATION

See the website at https://ee-workshop.for.lrz.de/