The Research Lab Computational Technologies and Applications (RL CTA) is since 2011 part of the unit «Theory and Applications of Algorithms (TAA)» at the «Faculty of Computer Science» (Fakultät für Informatik) of the «University of Vienna», Austria. It conducts research in computational technologies, including (but not limited to) the context of computational science research (as, for example, defined in a «PITAC report»).

 

A particular focus of the RL CTA is on numerical computations, investigating high performance aspects as well as the interaction of algorithms with state-of-the-art and future generation computer hardware. Concurrency and parallelism in contemporary computer architectures are dramatically increased at various levels to further improve the theoretical peak performance of computer systems. This widens the gap between the performance achieved with standard algorithms and the theoretical peak performance provided by state-of-the-art computer hardware. To narrow this gap, new algorithmic strategies need to be investigated which exploit concurrency at various levels of granularity as well as hardware accelerators, special purpose processors, etc. In this context, we investigate parallel and distributed algorithms for numerical matrix computations, concepts for controlled trading of accuracy against performance, as well as applications in computational science, data mining, machine learning and internet security.

 

The efforts of the RL CTA are partly funded by FWF, FFG, and by industrial partners, and its members are participating in multi- and interdisciplinary research projects with researchers from a variety of scientific fields.

FPGAs have evolved to powerful computing platforms. Today, multiple double-precision floating point units can be implemented in parallel, rivaling the performance of CPUs. In contrast to CPUs, however, architectures implemented on FPGAs can be custom-tailored to a specific task, thereby outperforming CPUs by an order of magnitude for selected kernels. Matrix multiplication is ubiquitous in scientific computing and its execution time is a key indicator for overall system performance.

 

You will learn how FPGAs can be used in high-performance computing, how to implement different architectures for matrix multiplication and how to write generic VHDL code. You will extend existing VHDL implementations and learn how to automate Altera Quartus II and to distribute the work load on large servers to synthesize multiple design variants for different FPGAs.

 

 

The official announcement can be downloaded as PDF from «here».