About the GPU COE at the University of Cambridge
The University of Cambridge became Europe’s first GPU Center of Excellence in 2008. The GPU COE was awarded for Cambridge’s early adoption of GPU computing for the acceleration of scientific software across a wide range of disciplines. The status also reflects the University’s commitment to the dissemination and teaching of CUDA to the wider UK academic and industrial communities.
About the Cambridge Many-Core Group
At the time the GPU Center of Excellence was awarded, Cambridge was at the forefront of GPU computing in several fields, including computational fluid dynamics, cosmological data analysis and medical image processing. Since then, with the support of NVIDIA, many other research areas at the University have benefitted from the power of CUDA enabled GPUs. These include: stereo image processing; gravitational N-body simulations; 3D ultrasound processing; tomographic particle image velocimetry analysis; DNA sequencing acceleration. With so many University departments involved (Engineering, Physics, Astronomy, Computer Science, Biomedical Science) it became clear that a new virtual group was required to collate and share the University’s GPU computing activities – this is the role of the Cambridge Many-Core Group.
About the PIs
Graham Pullan is a Lecturer at the Whittle Laboratory in the Cambridge University Engineering Department. He is also a Fellow of Trinity Hall. His research is concerned with the analysis, design and development of turbines and compressors for both propulsion and power generation applications. Computational fluid dynamics is a key tool in this effort and GPU computing enables either a order of magnitude reduction in simulation times for the same hardware cost, or an equivalent increase in the fidelity of the analysis. This step change is extremely important for the capabilities of both research institutions and industry.
Graham obtained his PhD from Cambridge in 2001. From 2001-2008 he was the Rolls-Royce Turbines Research Fellow at the Whittle Laboratory. From 2008, Graham has been the Mitsubishi Heavy Industries Lecturer in Turbomachinery.
Dr. Paul Calleja is Director of High Performance Computing at Cambridge University, where he provides research computing services across all academic disciplines. Dr. Calleja obtained his Ph.D. in computational bio-physics at Bath University. After filling a post-doctoral research position at Birkbeck College, he moved into private industry, where he spearheaded early commercialization of HPC cluster solutions within the U.K.
Following six years in the commercial sector, where he led the market transition from proprietary SMP systems to commodity cluster-based solutions. Dr. Calleja returned to academia to lead the formation of a new HPC service at Imperial College, London. From there, he moved to Cambridge University, to form a new HPC service and to direct a major reorganization that has resulted in University-wide HPC capabilities with a novel pay per use cloud computing model. Cambridge University now boasts one of the largest academic supercomputer in the U.K., occupying 20th position among the top 500 when first installed. Dr. Calleja sits on numerous national and international HPC committees and advisory boards, as well being a founding member of the U.K. HPC Special Interest Group.