NVIDIA NAMES BRAZIL’S FIRST CUDA CENTER OF EXCELLENCE

by Chandra Cheij

In another landmark for the CUDA parallel programming environment, NVIDIA has named the Fluminense Federal University (UFF) of Brazil as the country’s first CUDA Center of Excellence.

The recognition is a result of UFF’s ongoing work in parallel computing research and education using NVIDIA GPUs and the CUDA parallel computing environment.

Located in the state of Rio de Janeiro, UFF is one of Brazil’s top computer science research and education centers. Its departments have a large computing infrastructure, including a 32 Fermi GPU-based cluster, a virtual reality and simulation center and an extensive CUDA-enabled teaching lab.

UFF Named Brazil First CUDA Center of Excellence
UFF Named Brazil’s First CUDA
Center of Excellence

As one of the world’s 20 CUDA Centers of Excellence, UFF will use equipment and grants provided by NVIDIA to support research and academic programs across its mathematics, physical and life sciences departments.

“With NVIDIA’s support, we can continue to enhance our CUDA evangelism, engage in other CUDA-based research projects and develop new programs to reach larger numbers of researchers and students,” said Professor Esteban Clua, principal investigator at Fluminense Federal University

In one ongoing project, the Computer Science Department of UFF, together with Brazil’s National Institute for Space Research, is investigating new and alternative physics and computational models, with important results for modern physics. Requiring the computation of the interactions of billions of particles, the research is being implemented on a cluster made entirely of GPUs. According to the researchers, without GPUs it would take many years to achieve and validate the results that now take a few months.

This same group is also investigating and proposing a national GPU computing grid for cosmological research. Other important research projects involve wave propagation calculation for oil and gas exploration, crowd simulations, GPU and NVIDIA OptiX real-time ray tracing, and fluid simulations.

“GPUs and CUDA are making possible the dream of personal clusters, and this is drastically changing the way of conducting science,” said Clua.

The CUDA Center of Excellence program was created to foster collaboration with leading institutions at the forefront of parallel computing research and education.