¹û¶³Ó°Ôº

Since 1999, the EPSRC has been conducting a series of reviews of the state of UK science. The international review panel benchmark the strength of UK research activity against world competitors, and highlight any gaps or missed opportunities.

Each discipline is scheduled for review every five years, but this was the first time high performance computing (HPC) has been reviewed. In September 2005, a panel of computational scientists spent a week visiting 15 research groups at universities across the country, talking to researchers and students, discussing their observations, and formulating their recommendations. They judged that in many of the consortia and research groups visited, the scientific results compared well with the highest international standards.

The report states that '…several examples of the fruitful delivery of science through the use of HPC in the UK stood out.' These outstanding examples included Professor David Price's (¹û¶³Ó°Ôº Earth Sciences) Crystallography & Mineral Physics Group, who have carried out ab initio simulations on high pressure and temperature solid and liquid iron and its alloys, to determine the viscosity at the Earth's outer core, the high-pressure melting point of iron, and the probable composition of Earth's inner and outer core. The results were published in the journal 'Nature'.

Professor Richard Catlow (¹û¶³Ó°Ôº Chemistry) has pioneered the development and application of computer modelling in solid state and materials chemistry. Of particular note is his highly significant and widely quoted work on silicate minerals. His group was also commended: 'In the area of HPC simulations of materials and nanoscience, Richard Catlow's group has achieved major progress in studies of heterogeneous catalytic processes. In the course of their development of the most advanced molecular dynamics code, simulations of several million atoms have been carried out - a world record for simulations of materials with charged ions.'

The panel also stated: 'In our review of UK research using HPC, the panel found several examples of outstanding and productive collaborations between applications scientists and computer scientists or mathematicians. The Reality Grid project at ¹û¶³Ó°Ôº demonstrated some first-rate examples of computational steering that required Grid middleware, understanding of networking technology, and advanced visualisation.' Reality Grid is led by Professor Peter Coveney (¹û¶³Ó°Ôº Chemistry), who is also co-Director of the ¹û¶³Ó°Ôº e-Science Centre of Excellence.

• Links:
  
•