Logo des Repositoriums

Auflistung nach Autor:in "Mier, Eric"

1 - 2 von 2
  • Konferenzbeitrag
    Energy Elasticity on Heterogeneous Hardware using Adaptive Resource Reconfiguration
    (Datenbanksysteme für Business, Technologie und Web (BTW 2017), 2017) Ungethüm, Annett; Kissinger, Thomas; Mentzel, Willi-Wolfram; Mier, Eric; Habich, Dirk; Lehner, Wolfgang
    Energy awareness of database systems has emerged as a critical research topic, because energy consumption is becoming a major factor. Recent energy-related hardware developments tend towards o ering more and more configuration opportunities for the software to control its own energy-based behavior. Existing research within the DB community so far mainly focused on leveraging this configuration spectrum to identify the most energy-efficient configuration for specific operators or entire queries. In [Un16], we introduced the concept of energy elasticity and proposed the energy-control loop as an implementation of this concept. Energy elasticity refers to the ability of software to behave energy-proportional and energy-e cient at the same time while maintaining a certain quality of service.
  • Konferenzbeitrag
    Overview on Hardware Optimizations for Database Engines
    (Datenbanksysteme für Business, Technologie und Web (BTW 2017), 2017) Ungethüm, Annett; Habich, Dirk; Karnagel, Tomas; Haas, Sebastian; Mier, Eric; Fettweis, Gerhard; Lehner, Wolfgang
    The key objective of database systems is to e ciently manage an always increasing amount of data. Thereby, a high query throughput and a low query latency are core requirements. To satisfy these requirements, database engines are highly adapted to the given hardware by using all features of modern processors. Apart from this software optimization, even tailor-made processing circuits running on FGPAs are built to run mostly stateless query plans with a high throughput. A similar approach, which was already investigated three decades ago, is to build customized hardware like a database processor. Tailor-made hardware allows to achieve performance numbers that cannot be reached with software running on general-purpose CPUs, while at the same time, addressing the dark silicon problem. The main disadvantage of custom hardware is the high development cost that comes with designing and verifying a new processor, as well as building respective drivers and the software stack. However, there is actually no need to build a fully-fledged processor from scratch. In this paper, we present our conducted as well as our ongoing research e orts in the direction of customizing hardware for databases. In detail, we illustrate the potential of instruction set extensions of processors as well as of optimizing memory access by o oading logic to the main memory controller.