Analyzing the Energy Behavior of Relational Operators in Modern Database Systems
Master's Thesis
Christian Matuschek
As power consumption of huge datacenters is growing rapidly, energy efficiency becomes more and more important. At the same time, modeling and understanding the CPUs has become much more challenging. Therefore, we try to provide a tool for empirical analysis of modern CPUs processing database operations.
The free DBMS software MonetDB will be used as a framework for the implementation. This tool provides an in-depth view of the database sub-operations and their energy and memory behavior. Further, we analyze a realistic scenario, to provide insight of potential energy efficiency in database operations.