"Computer science also differs from physics in that it is not actually a science. It does not study natural objects. Neither is it, as you might think, mathematics; although it does use mathematical reasoning pretty extensively. Rather, computer science is like engineering - it is all about getting something to do something, rather than just dealing with abstractions as in the pre-Smith geology. "

Richard Feynman 

(from the Feynman Lectures on Computation)




Tony Hey

Professor of Computation in the Department of Electronics and Computer Science, University of Southampton

Seconded as Director of UK e-Science Core Programme (from April 2001), EPSRC, Swindon


Tony Hey joined the ECS Department in 1986 and founded Concurrent Computation Research Group with colleagues Denis Nicole, David Pritchard and Jeff Reeve. Prior to his move to ECS he was a Reader in Theoretical Physics at Southampton working on computational particle physics. His interests in parallel computing date back to a sabbatical in Caltech in 1981 where a seminar by Carver Mead demonstrated that there were no engineering obstacles to Moore's Law continuing for the next twenty years or so. The implication of this resulting exponential growth of both microprocessor performance and memory density for computational science was that simulation of realistic large-scale systems would be possible on parallel assemblies of VLSI chips. 

On his return from Caltech he explored the use of the ICL DAP 4096 single bit processor SIMD machine for scientific simulations. In many ways the DAP and its array parallel DAP Fortran language was a precursor of the Connection Machine. On visits to Caltech in the early 1980's  he explored MIMD computing on Fox and Seitz's Cosmic Cube. In ECS, with colleagues Chris Jesshope and Denis Nicole, he initiated the EU-funded Esprit SuperNode project with Inmos to design and build reconfigurable networks of new 'floating point' T800 transputers developed in the project. After the success of the Supernode project followed by the comparative failure of the T9000 transputer, his research interests centred on the problem of parallel program portability on parallel systems built from mainstream commodity components. 

With Jack Dongarra, Rolf Hempel and David Walker, he wrote the first draft of a specification for a new message-passing standard called MPI [C55]. This initiated the process that led to the successful MPI standard of today. He led the performance evaluation Work Package in the Esprit and developed the first set of Distributed Memory message-passing parallel benchmarks. Following on from this, with Roger Hockney and Jack Dongarra he was instrumental in setting up the international Parkbench Initiative for Parallel Kernels and Benchmarks. With a series of students at Southampton, he has investigated a variety of techniques for parallel performance estimation ranging from 'white box' benchmarking [C32] to the PERFORM method for rapid simulation [C30]. His research interests continue to be centred around 'performance engineering' now extended to the context of the Grid. From his experience in writing up Richard Feynman's 'Lectures on Computation' he retains a keen interest in  Quantum Computing.  

Tony Hey also has a passionate interest in communicating the excitement of science to young people. He has written 'popular' books on quantum mechanics [U1, U3] and on relativity [U2 ]. He is now contemplating a popular book on computing. He has also been Head of Department and Dean of Engineering at Southampton.

Tel +44 (0)1793 444022
Fax +44 (0)1793 444456

Postal Address:
Prof A J G Hey


Polaris House

North Star Avenue