Motivation

Energy efficiency has become a decisive factor in modern computing systems. For multiple reasons, many of today's systems are effectively limited by their energy consumption:

• In battery-powered devices high energy consumption will quickly drain batteries.
• In large data centers energy effectively has to be invested twice: energy consumed by computing machines will be dissipated as heat; removing this heat (through cooling and air conditioning systems) again requires energy.
• Modern processor designs are, in fact, all energy-limited. Moore's law would, in theory, offer an exponentially growing amount of compute ressources (transistors) to chip designers. However, actually operating all these ressources would lead to excessive heat that could not be dissipated through typical chip packages.

Project

The behavior of algorithms and programs may have important effects on the energy consumption characteristics of a machine. As part of the sub-project A2: Algorithmic Aspects of Learning Methods in Embedded Systems of SFB 876, we investigate how we can improve energy efficiency by making data-intensive algorithms more energy aware.

People

• Jens Teubner
• Henning Funke

This is a joint project with the Efficient Algorithms and Complexity Theory Group at TU Dortmund.

Publications

• Henning Funke, Jan Mühlig, and Jens Teubner. Efficient Generation of Machine Code for Query Compilers. 16th Int'l Workshop on Data Management on New Hardware (DaMoN), Portland, OR, USA, June 2020.
• Henning Funke, and Jens Teubner. Data-Parallel Query Processing on Non-Uniform Data. Proceedings of the VLDB Endowment, vol. 13(6), February 2020.
• Stefan Noll, Jens Teubner, Norman May, and Alexander Böhm. Accelerating Concurrent Workloads with CPU Cache Partitioning. Proc. of the 34th Int'l Conference on Data Engineering (ICDE 2018), Paris, France, April 2018.
• Henning Funke, Sebastian Breß, Stefan Noll, Volker Markl, and Jens Teubner. Pipelined Query Processing in Coprocessor Environments. Proc. of the 2018 ACM SIGMOD Conference on Management of Data, Houston, TX, USA, June 2018.
• Stefan Noll, Henning Funke, and Jens Teubner. Energy Efficiency in Main-Memory Databases. Datenbank-Spektrum, Springer Verlag, July 2017 (Online First).
• Stefan Noll. Energy Efficiency in Main-Memory Databases. BTW 2017.