The IDEaS Seminar Series and the Center for High-Performance Computing (CHiPC) present Dr. Sadaf Alam of the Swiss National Supercomputing Centre (CSCS)
"Evolving an HPC Center for Diverse Workloads and Workflows"Friday, September 1, 2017, from 2-3 pm, in the TSRB Auditorium
Supercomputing platforms and their hardware/software ecosystems at high-performance computing (HPC) data centers have been designed and tuned over the years to serve the needs of cutting edge simulation-based science. However, the emergence and rapid growth of data science problems, solutions, and associated technologies have created new demand for computing resources for an even broader range of business and scientific research communities. As a result, many computing centers are asking how to "pivot" their hardware and software services to fulfill the needs of increasingly diverse user communities without duplicating or multiplying resources. For instance, a common question is how to use co-processor technologies, like GPUs, for both simulation science and data science workloads, given their often differing software stacks. In this talk, I will give an overview of the evolution of technologies for simulation and data science workloads and workflows. I will compare and contrast technical requirements using two case studies, one representing each domain. I will then highlight challenges and opportunities for using consolidated technologies. Finally, I will discuss ongoing research and development opportunities that could offer a significant improvement in efficiency for diverse workloads and complex workflows by using converged solutions.
Dr. Sadaf R. Alam is the Chief Architect and Head of the HPC Operations at the Swiss National Supercomputing Centre (CSCS) in Lugano, Switzerland. Dr. Alam studied computer science at the University of Edinburgh, UK, where she received her Ph.D. in 2004. Until March 2009 she was a computer scientist at the Oak Ridge National Laboratory, USA. At CSCS, Dr. Alam leads the unit operating HPC systems for the User Lab and external customers, like MeteoSwiss, the Federal Office for Meteorology and Climatology. In her role as Chief Architect, she ensures the end-to-end integrity of HPC systems and storage solutions. Her research interests include the development of tools and technologies for emerging computing, memory, network and storage technologies.
Richard Vuduc, Associate Professor, School of Computational Science and Engineering at Georgia Tech; Director, Center for High-Performance Computing at Georgia Tech.