Energy Efficiency in Computing Centres: Parallel session A3
- Mike Ashworth (STFC)
Prof. Thomas C. Schulthess (CSCS)
High-performance computing (HPC) data centers are here to deliver simulation capabilities as tools for scientific discovery. Supercomputers comprise the most capital-intensive part of this infrastructure and are thus receiving most of the attention. However, with power consumption of these systems jumping from the typical hundred kilowatts a decade ago to several megawatts in recent years, and...
29. Exploiting mobile phone technology to build energy efficient supercomputers: the Mont Blanc project
Dr Simon McIntosh-Smith (University of Bristol)
Energy efficiency is becoming the limiting factor in large-scale high performance computing systems. Today’s fastest machines already use megawatts of power and cost millions of dollars a year to run. By the end of this decade, the fastest machines are predicted to require over 20 megawatts of power, making them considerably more expensive to run than today’s machines, and limiting the number...
Dr Bruno Michel (IBM)
Chip microscale liquid-cooling reduces thermal resistance and improves datacenter efficiency with higher coolant temperatures by eliminating chillers and allowing thermal energy re-use in cold climates. This concept has been successfully demonstrated three years ago in a one rack prototype and scaled up to a 10'000 server system. Liquid cooling enables an unprecedented density in future...
Mr Wayne Salter (CERN)
Over the past few years, CERN has been working to improve the efficiency of its main Data Centre that was built in the early 70s. Most of the equipment hosted in the Centre relies on an air-cooling infrastructure that was designed and installed many years ago. By applying small changes to the existing, and ageing, cooling infrastructure, CERN has significantly improved the energy efficiency...