The HPDC Cluster conducts pioneering research in software for massively parallel computing systems. Research considers systems at all scales, from custom heterogeneous many-core designs for the embedded systems domain to Exascale systems and Datacentres.
HPDC adopts a holistic approach that addresses simultaneously the challenges of high performance, performance portability and dependability, energy-efficiency and reliability. Ongoing projects build upon and explore disruptive hardware technologies, developing methods to seamlessly integrate these technologies into the high-performance computing ecosystem and, in the long term, the mainstream computing ecosystem.
HPDC has expertise in all layers of the system software stack, including languages, libraries, compilers, runtime systems, operating systems, virtualisation environments, firmware and tools. Expertise in system software is complemented by expertise in theoretical foundations of parallel computation, where Queen's University has a long-standing tradition, computational science and high-performance algorithmics.
Research projects in HPDC are largely application-driven. The Cluster regularly interacts with industry to revise its research and education portfolio, towards training a next generation of experts who are ideally positioned to contribute HPDC technology to emerging market domains. As part of this effort, HPDC currently explores research, development and training activities in the domains of High Performance Computational Finance and Real-Time Big Data Analytics.
HPDC has an extensive network of collaborators, ranging from leading technology providers such as IBM, Intel, ARM, SAP and HP, to SMEs, and from supercomputing centres such as EPCC, LLNL, ORNL, BSC, JSC and HLRS to peer academic groups. Many of these partnerships materialise in collaborative international projects funded by EC Framework Programmes, EPSRC and the industry. The Cluster has active memberships and significant involvement in the European Network of Excellence on High Performance and Embedded Architectures and Compilers (HiPEAC) and the EPSRC Network on Numerical Algorithms and High Performance Computing.