Parallel Algorithmics
This page is obsolete! Please go to
the new department website.
Parallel Algorithmics
Contact person: Marian Vajteršic
Description
The research activities of this working group focus on the design and implementation of efficient parallel algorithms for solving relevant theoretical and practical problems in computer science.
The goal of designing these algorithms is to reflect the architecture and software characteristics of current supercomputers in order to exploit their performance potential to a maximum extent.
The research themes include:
- Algorithms for retrieval on large data collections by algebraic and clustering methods
- Architecture and software design of a parallel automatic data retrieval systems
- Parallel linear algebra methods (matrix-multiplication, systems of linear algebraic equations, eigen- and singular value solvers)
- VLSI solvers for elliptic partial differential equations
- GPU-based optimization for medical image processing applications
- Signal processing and image and video compression algorithms for media servers and media processors.
- Parallelization of spiking neural networks and evolutionary strategies on hybrid clusters
Our algorithms are implemented under consideration of modern parallel programming paradigms, using parallel computer systems installed at the Department of Computer Sciences of the University of Salzburg () and supercomputers of collaborating partner institutions abroad.
