We conduct research in the development of new hardware and software architectures for high performance computing using a variety of technologies including Field Programmable Gate Arrays (FPGAs), Graphics Processing Units (GPUs), and multi-core processors. The aim is to deliver orders of magnitude speed-ups over existing solutions at reasonable costs. In particular, increased power consumption in traditional high performance computing platforms is opening a window of opportunity for low power hardware accelerators including FPGAs. The Maxwell FPGA supercomputer at the University of Edinburgh is one of few multi-FPGA systems constructed for this purpose worldwide.

We collaborate closely with colleagues at the FPGA High Performance Computing Alliance (FHPCA) with partners at the Edinburgh Parallel Computing Centre, AlphaData, Nallatech and other start-ups and SMEs in Scotland and beyond, with the aim of bridging the gap between application developers in the scientific computing community and low level FPGA hardware.

Current research projects in this area include the development of efficient FPGA-based implementations for bioinformatics and computational biology applications e.g. biological sequence analysis and molecular dynamics simulation, financial computing e.g. option pricing, and large data mining.