Senior Computational Scientist - Multiscale Workflows

Senior Computational Scientist- Multiscale Workflows

Science and Technology Facilities Council

Salary: £39,748 to £44,166 (dependent on qualifications and experience)

Plus a recruitment and retention allowance of up to £4,000 per annum

Contract Type: FTC (32 months)

Hours: Full time/Part Time (minimum of 25 hours a week)

Location: STFC, Daresbury Laboratory, Cheshire

Closing date: 27th February 2022

We have an exciting vacancy for a Senior Computational Scientist in the Scientific Computing Department (SCD) at STFC, to join the team developing modern Materials and Molecular Modelling packages on High Performance Computing (HPC) platforms. Your work will focus on multiscale modelling algorithms and complex workflows.

The work forms part of the recently funded Particles At eXascale on High Performance Computer (PAX-HPC) grant (EP/W026775/1) funded under the UKRI ExCALIBUR (Exascale Computing ALgorithms & Infrastructures Benefiting UK Research - https://excalibur.ac.uk/), a UK research programme that aims to deliver the next generation of high-performance simulation software for the highest-priority fields in UK research.

PAX-HPC is a £3M project to port and optimise high impact particle-based scientific codes to modern HPC architectures. For the first time it brings together experts in Materials and Molecular Modelling (MMM) and Smoothed-Particle Hydrodynamics (SPH), and is a large collaboration within the UK including UCL, Durham, Manchester, York, STFC, Lincoln and Southampton amongst other centres. Within the project a total of 7 RSEs are funded, and as a central strand of the grant is knowledge exchange the post advertised here will form part of a large community over the coming years.

Work Programme

Your research will focus on complex parallel workflows with an emphasis on first principles multiscale modelling simulations, driven by the needs of applications in catalysis and electrochemistry. The work will involve researching and implementing advanced QM/MM approaches involving multiple QM regions in the multiscale modelling package ChemShell, and investigations into new MMM workflows for exascale platforms in close collaboration with our PAX-HPC partners, including GPU-aware execution, hybrid algorithms interfacing to leading MMM software developed by our partners (including CP2K, QUESTAAL, CASTEP and CRYSTAL), and integration with exascale-ready C++ codes and workflow management systems. A second strand of research will target exascale I/O applications, through a domain-specific API to middleware such as HDF5 and ADIOS2, to be showcased in DL_POLY, QUESTAAL and CASTEP.

As noted above a central tenet of the grant is knowledge exchange. As such you will attend regular meetings with the other RSEs within PAX-HPC, and will also present your work at larger meetings of the community. Further you will be encouraged to interact where appropriate with other projects funded under the ExCALIBUR umbrella, and with the UK RSE community in general.

Contacts and Communication

The post will involve extensive team-working both within the group and with partner organisations.

You will work with Dr Thomas Keal in the Chemistry Group at STFC Daresbury Laboratory, working together with Dr Alin Elena, and in collaboration with a number of other institutions, especially Prof Scott Woodley and Dr Alexey Sokol at University College London, Dr Phil Hasnip at York University and Dr Dimitar Pashov at King’s College London.

The Scientific Computing Department of STFC has an excellent track record in high end computing, scientific software development and related research. The group has access to a number of STFC HPC platforms, as well as access to the ARCHER national supercomputer and some Tier 2 HPC centres.

For further information about this position please contact Thomas Keal (Email: thomas.keal@stfc.ac.uk).

About you

The successful applicant will have, or about to be awarded, a PhD and ideally a good track record of research in a computational science discipline, or equivalent experience. A background in computational MMM research and the theory that underpins it is desirable, but not essential. Knowledge of modern HPC architectures is essential.

Candidates must be able to demonstrate strong software engineering skills, with a track record of significant contributions to scientific software projects. They will be comfortable with working with compiled and interpreted programming languages and have a good understanding of modern high performance computing. Practical experience of parallel programming techniques will be essential. Fluency in any of the main languages employed in the projects (Fortran 2003, C, CUDA and Python) is particularly desirable. Contributors to large scale scientific packages, especially those that fit under the MMM remit, are encouraged to apply.

The post will include regular travel to project partners, and occasional travel to scientific conferences in the UK and overseas.

Essential:

• A PhD (or equivalent experience) in a physical science or other relevant discipline
• A demonstrable track record of software engineering
• Knowledge of at least one of modern Fortran, C or C++
• Knowledge of a scripting language (e.g. Python)
• Experience in a Unix/Linux command line environment
• Practical experience of at least one parallel programming paradigm (e.g. MPI, OpenMP, GPUs)
• A track record of working as part of a team
• Good written communication skills as exemplified by technical reports, peer reviewed papers and/or similar material (e.g. a dissertation as part of a postgraduate degree)
• Provision of a piece of relevant code that the candidate has written.

Desirable:

• Experience of working with Materials and Molecular Modelling codes, especially MD or local basis set ab initio codes (for instance DL_POLY, Tinker, CP2K, CRYSTAL)
• Experience of at least one of
  • GPU programming
  • developing and implementing complex workflows
  • multiscale modelling (e.g. ChemShell)
  • knowledge of parallel I/O frameworks, e.g. HDF5 or ADIOS2
• Demonstrable contributions to scientific software packages
• Experience with software engineering tools, such as revision control, debuggers and profilers
• Experience of using a high performance computing facility
• Familiarity with mixed language programming.

What we can offer you

By choosing to come and work at the STFC part of UKRI means that you will have access to a whole host of attractive benefits!

These include:

• An exceptional index linked pension scheme,
• 30 days leave (pro-rata) plus 10 public holidays,
• Christmas shut down
• Flexible working,
• Access to employee shopping/travel discounts
• Salary sacrifice cycle to work scheme.
• Social and sporting activities and societies
• We are committed to developing our staff, and training will be provided in relevant areas.

We are an equal opportunity employer and value diversity at our company.