ReProVide (Query-Optimierung und Near-Data-Processing auf rekonfigurierbaren SoCs für Big Data Analyse)This project is conducted together with the Chair for Computer Science 12 (Hardware-Software Co-Design, https://www.cs12.tf.fau.eu/ ). They provide a platform for the fast execution of queries with the help of dynamically partial reconfiguration of FPGAs. Our contribution is to include the use of this platform into the query execution wherever this has an advantage. That can be in addition to acceleration of query execution also the saving of energy. For that purpose, the query optimizer must be extended. It must know about the platform and its potential, e.g. as a data manager of its own with special operators, and must then use it if this is advantageous for a query. The cost models of the query optimization must therefore be adapted in an appropriate way. | Project manager: Prof. i. R. Dr. Klaus Meyer-Wegener
Project participants: Dr.-Ing. Lekshmi B.G.
Keywords: Query optimization; cost models; hardware accelerators
Duration: 1.5.2018 - 30.4.2021
Sponsored by: DFG
Mitwirkende Institutionen: Lehrstuhl für Informatik 12
Contact: B.G., Lekshmi E-Mail: lekshmi.bg.nair@fau.de
| Publications |
---|
Ziener, Daniel ; Bauer, Florian ; Becher, Andreas ; Dennl, Christopher ; Meyer-Wegener, Klaus ; Schürfeld, Ute ; Teich, Jürgen ; Vogt, Jörg-Stephan ; Weber, Helmut: FPGA-Based Dynamically Reconfigurable SQL Query Processing. In: ACM (Ed.) : ACM Transactions on Reconfigurable Technology and Systems (TRETS). Vol. 9(4). New York, NY, USA : ACM Press, 2016, pp 25:1-25:24.ISSN 1936-7406 [doi>10.1145/2845087] |
|