We are computational scientists and specialists research software engineers who combine science domain knowledge with software engineering skills. We care about building and maintaining knowledge, expertise, and software consistently and sustainably – and have done so for over 40 years. Our scientific expertise spans a range of lengthscales in computational science and engineering. We are organised in the following groups:
The theoretical and computational physics group (led by Dr Gilberto Teobaldi), specialises in the development and application of theoretical and computational approaches, mainly from first principles, for the simulation of static and dynamics properties of atoms, molecules and materials and their response to external stimuli (e.g. light, temperature, pressure, etc) as well as magnetically confined plasmas.
The computational chemistry group (led by Prof Ilian Todorov), specialised in the utilisation and advancement of classical methodologies for molecular simulation and HPC software engineering techniques. A large part of our specialism is incorporated in a suite of software packages, known as DL_Software, bridging the atomistic scale with the quantum and the mesoscopic. In addition to statistical physics and thermodynamics, the group’s wider expertise also includes DFT, minimisation techniques, advanced sampling, Dissipative Particle Dynamics and Lattice Boltzmann.
The computational biology group (led by Dr Martyn Winn), develops and applies methods for computational structural biology, tomography, molecular simulation and genomics. The group has expertise in diffraction physics, volume reconstruction, classical molecular dynamics, sequence analysis, machine learning, etc.
The computational engineering group (led by Prof David Emerson), focused on general areas of Computational Fluid Dynamics covering nuclear engineering, thermohydraulics, rarefied gas dynamics, lattice Boltzmann method, smoothed particle hydrodynamics, turbulence and acoustics. Major emphasis is placed on HPC covering Peta/Exascale application development and on multiscale/multiphysics applications related to creating “digital twins”.
(led by Dr Luke Mason)
focuses on providing the CCPs and High End Computing (HEC) Consortia with the knowledge and skills needed to improve their world-leading scientific software. The needs of these communities were assessed using a very successful software audit in early 2018. This audit was a vital component in ensuring that the delivered work will have maximum impact within these supported communities, thereby allowing them to reach the needs of their communities and utilise current and near-future computer architectures.
We achieve this by:
- evaluating new programming tools, libraries and techniques that are essential for the timely and cost-effective exploitation of current and near-future systems;
- demonstrating how specific methodologies can be applied to existing applications;
- developing comprehensive (online) training packages, which can also be delivered in person.
Recent activities have included:
- evaluation and provision of training in the use of reduced precision in computational simulations to both speed-up codes and reduce the energy consumed when running these codes;
- investigation into the scalability of coupled problems that are being developed by CCP-WSI;
- evaluation of and provision of training in the use of the open source profiler TAU to usefully profile complex simulation codes.
Mid to long-term outcomes:
- Comprehensive training provisions to suit the needs of the CCPs/HEC Consortia
- Comprehensive evaluations of programming tools and methodologies to suit the needs of the CCPs/HEC Consortia