hpc-ch
  • Facebook
  • Twitter
  • Youtube
  • Rss
The Swiss HPC Service Provider Community
  • Home
    • Calls for Proposals
    • Conferences & Presentations
      • Video of talks
    • Courses & Workshops
      • Video of Courses
    • Job offers
    • Science
    • Technology
    • Video blog
  • About
    • hpc-ch Community News
    • hpc-ch Booth
    • Forums
    • HPC Advisory Council
  • Members Academia
    • [BC]2
    • CHIPP
    • CSCS
    • EMPA
    • EPF Lausanne
    • ETH Zurich
    • PSI – Paul Scherrer Institut
    • Speedup
    • SwiNG
    • SWITCH
    • SystemsX.ch
    • Università d. Svizzera italiana
    • Universität Basel
    • Universität Bern
    • Université de Fribourg
    • Université de Genève
    • Université de Lausanne
    • Universität Zürich
    • Vital-IT
    • WSL
  • Members Industry
    • Casale Group
    • Credit Suisse
    • Hilti
    • MeteoSwiss
    • Novartis
    • PartnerRe
    • Syngenta Crop Protection
  • Contact
Home» Members Academia » ETH Zurich » Cost and Energy Analysis of Server Rooms at ETH Zurich

Cost and Energy Analysis of Server Rooms at ETH Zurich

Posted on November 29, 2010 by mdl in ETH Zurich, Technology

We are glad to share with you the results of the Master Thesis (in German) of Christoph Möder entitled «Suggestions for optimizing the server room at the ETH Zurich based on a cost and energy analysis». Providers of HPC services are increasingly faced with questions about energy costs and this study provides a good reference point.

You can download the main document or the appendix of the master thesis as PDF.

In his study, Mr. Mäder analyses first the power consumption of the existing IT servers and his infrastructure as their total cost of ownership (without building costs). The share of electricity consumption in the server rooms (excluding CSCS) is 25% of total electricity consumption of the ETH Zurich (page 27). The total cost (excluding CSCS) is  27 million CHF per year (page 29).

The breakdown of electricity consumption by application and the cost structure are shown for individual data centers and server rooms in Chapter 4.2 and 4.4.

In Figure 21 (page 28) the author computes for each data center the energy density and the PUE values. Figure 23 (page 29) shows also the cost per server room space and the cost per IT service.

In Chapter 6 (page 33), Mr Maeder gives suggestions and recommendations how the energy consumption and costs can be optimized. In Figure 31 (Page 4) the savings potential of these optimization approaches, depending on cost and feasibility are presented.

Bookmark and Share
ETH Zurich, Infrastructure, Power Consuption

Featured video

Discussion between Daniel Duffy and William Putman (NASA): Challenges for global climate simulation

Latest hpc-ch Tweets

  • Slidecast (in Italian): Grid computing and the search for the new particle at CERN by Günther Dissertori, ETH Zurich http://t.co/iNt5eyX74F
    May 16, 2013
  • Special ISC’13 Session to Probe the Thinking behind Europe’s Human Brain Project #epfl http://t.co/k8ZyuQe3dF
    May 13, 2013
  • CSCS Call for Proposals – Allocation period starting on 1 October 2013 http://t.co/Net2H0YCvq
    May 6, 2013

Posts by Category

(c) 2013 www.hpc-ch.org