Prof Ian Watson (BSc PhD FBCS CEng CITP) - research
My major research interest is in Parallel Computer Systems with particular emphasis on achieving high performance from parallel systems with minimal effort.
This implies programming styles which avoid explicit statements concerning parallelism wherever possible. This includes parallel implementations of declarative languages but not to the exclusion of more pragmatic approaches.
My expertise is really in machine architecture and run-time systems rather than language design and implementation, although I am convinced that a wide knowledge ranging from programming models through to hardware design is necessary to make progress in the area.
Recently, the processor industry has moved towards multi-core systems as the performance of single processor designs has become limited by a number of technological factors. Parallel programming, which was once the preserve of specialist programmers and limited applications, now needs to become both mainstream and general purpose.
Because of the limits imposed by technology, industry is looking to the future where it will be necessary to use hundreds or thousands of processors cores on a single integrated circuit in order to achieve higher performance. At this point, current ways of building and programming multi-core processors will simply not be viable.
My current research is therefore focussing on these issues. New computer architectures are being investigated which will allow extensible parallel structures. This is coupled with new programming models and programming languages to overcome the complexity of current approaches.