Announced in June 2016, CFMS awarded the contract for the refresh of its high performance computing infrastructure to Cray Inc., a global expert in supercomputing. The new system consists of a Cray CS400 cluster supercomputer, utilising Intel Xeon processor E5-2600 v4 product family formerly codenamed Broadwell with a parallel storage system provided by ArcaStream. Also included are NVIDIA Tesla Graphical Processing Unit (GPU) accelerators used to accelerate applications, which will provide customers with faster response times and enhanced performance levels for demanding computational tasks.
Up to five times more powerful than the previous system, industry organisations including Rolls-Royce and Airbus use the CFMS High Performance Computing (HPC) cluster for research and methods development, which underpins simulation-based design activities. Undertaking a collaborative role, CFMS acts as an interface between industry, academia and technology organisations, offering access to advanced simulation expertise and technologies that enable companies to accelerate product development and engineering design capability.
Sam Paice, Chief Operating Officer, stated: "With an evolution in engineering design taking place, we are welcoming a new era, one where advanced simulation takes centre stage, accelerating product development and differentiation, while enabling technological advantages and business growth. We are pleased to announce the availability of the new supercomputing cluster at CFMS."
"Cray is proud to be working with CFMS to deliver on its commitment to high value design capability. The CS400 system selected will provide the computational resources for CFMS's user community, which will lead to greater efficiencies in engineering design that Cray is privileged to be associated with", stated Dominik Ulmer, VP Business Operations EMEA at Cray.
CFMS operate across a number of sectors including aerospace and defence, automotive, civil engineering and renewable energy. Access to the new system is open to companies large and small.