Abstract : Power consumption is a critical factor for servers and data centers but do you know how much power your servers use and how much energy your applications consume, in which part of the server this power is consumed, and what can we do to reduce it?

After an introduction on the different sources of power consumption in a data center (power consumption, cooling and power loss), we’ll dig into the different components of the power consumption in a server when idle or when  running different workloads. We’ll show the relation between power consumption and performance on different processor generations and how we can take advantage of the trade-off between power and performance.

In a last part, I’ll present the Energy Aware Scheduling features we introduced in Platform LSF and the experience we had with similar features introduced since June 2012 on SuperMUC at LRZ with IBM Load Leveler. I’ll show how these features help control and optimize the power, performance and energy on a large cluster running many different parallel applications.

Biographie :Dr. Luigi BROCHARD Luigi has a Ph.D in Mathematics from Pierre et Marie Curie University in Paris, France and a HDR in Computer Sciences from the same university. Before joining IBM, he was assistant professor in Computational Sciences and worked as visiting scientist in different IBM labs. In 1990 he joined IBM France Europe introducing the IBM SP1 system and became the lead architect of the IBM HPC technical team in 2000. As such he was the architect of many large HPC systems IBM installed in Europe. From 2006 to 2008 Luigi worked in Austin, TX with the IBM HPC Development team where he created new methods to optimize power and energy consumptions of servers and clusters. This technology now called Energy Aware Scheduling has been included in IBM Tivoli Loadleveler and IBM Platform Load Sharing Facility (LSF). Back in Europe, Luigi lead the  team designing SuperMUC system at LRZ which was installed in 2012 as the largest and most energy efficient x86 system in the world. He is expert for European projects like EESI (European Exascale Software Initiative) and FP7 Infrastructure Projects. He is a member of the Green GRID and Energy Efficiency HPC Working group (EE HPC WG/LBNL) and author of many technical papers published in various international journals and conferences. In 2004, he was named IBM Distinguished Engineer and is now WW HPC Technical Director for IBM x86 and Pure Systems.