Cambridge University uses Dell's High Performance Computing

Dell and The University of Cambridge have teamed up to launch the Cambridge High Performance Computing (HPC).

The partnership between the two is said to be an effort to form the EMEA region’s top HPC community. This applies to both academic and commercial sectors.

The Centre is said to combine large-scale, commodity-based HPC infrastructure with experienced and specialised research, which is hoped to overcome the traditional barriers of entry to HPC.  Teams from Dell, The University of Cambridge and other vendors will build and test research-specific HPC tech, with the findings released through a series of free whitepapers and technical bulletins.

The University of Cambridge is of course no stranger to research and technology, claiming to  form the central hub of Europe’s largest technology centre with more than 1,200 technology companies located in science parks surrounding the city.

The Centre will be based out of the existing Cambridge HPC Service building, a facility already being used for delivering HPC services via a cloud computing model. On the University campus, the HPC cluster supports around 400 internal users spread across 70 research groups ranging from traditional hard sciences such as chemistry, physics and biology.

Dr. Paul Calleja, Director HPC Service, who is leading the project said: “We have amassed a considerable and focused pool of expertise and compute power that we hope will help speed up research within a wide range of fields. For researchers taking their first steps into HPC, we can provide the perfect platform to trial applications and for those looking to take things to a new level, we have the necessary support and understanding to really get research off the ground.”

The new centre will research lustre parallel file systems, storage design, implementation and operation as well as GPU Clusters – CUDA programming and GPU cluster system integration.

Other research will look at scientific visualisation, application optimisation and benchmarking, MPI profiling.