Research Software Engineer

Women in HPC; Town of Cambridge; On-site; AUD 34,000 - 44,000; Part time

Job summary

The University of Cambridge invites applications for a part-time Research Software Engineer (0.5 FTE) to join the Institute of Astronomy and Kavli Institute for Cosmology. You will develop and maintain GPU-accelerated HPC software, implement testing and documentation, and coordinate shared code with partner institutions.

Work involves engaging in modern software engineering practices, contributing to open-source workflows and potential publications, with growth opportunities as the Cambridge

Qualifications

• Experience developing high-performance scientific software.
• Strong programming in C++ and Python for HPC workloads.
• Familiarity with GPU acceleration and GPU-aware tools.
• Ability to document, test and maintain codebases for collaboration.

Responsibilities

• Establish and manage reliable version-control infrastructure for group codes with migration of legacy repos.
• Coordinate external development branches and integrate selected changes into group codes.
• Tune and deploy codes on HPC clusters (AMD/NVIDIA GPUs) with profiling, benchmarking and scaling.
• Design, implement and document new physics capabilities within group codes with performance in mind.
• Prototype ML and advanced statistics methods to speed modeling and inference workflows.
• Support development of semi-analytic modelling codes and integrate into open-source frameworks.
• Implement automated testing, CI and documentation across codebases; provide training.
• Advise group leader on software strategy and compute-resource planning for grants.
• Contribute to publications and pursue new computational ideas for group projects.

Skills

C++

Python

GPU CUDA

HPC

Software engineering

Version control

CI/CD

Documentation

Bayesian methods

Machine learning

Tools

CUDA

HIP

OpenCL

Git

CI/CD tooling

Job description

Fixed-term: The funds for this post are available for 5 years in the first instance. If additional funding is acquired, extension may be possible.

The Institute of Astronomy (IoA) and the co-located, interdepartmental Kavli Institute for Cosmology, Cambridge (KICC) at Cambridge University are internationally renowned for a uniquely stimulating and multi-disciplinary research environment. One topic where Cambridge has a long history is black holes, the most extreme environments in nature, and laboratories where strong gravity, plasma physics and high-energy astro(particle)/physical processes meet. The research group of Prof. Sera Markoff (Plumian Professor) studies these systems across the full black hole mass scale, combining GPU-accelerated supercomputer simulations and semi-analytic modelling with data from world-leading facilities including the Event Horizon Telescope (EHT). Her group sits at the nexus of several international efforts: a growing Cambridge node, where this post is based (currently expanding over the coming years into a fully-fledged team of students and postdocs); an established group at the University of Amsterdam of several postdocs and students; her ongoing ERC Synergy Grant project ‘Blackholistic’, with partner groups at Oxford, Radboud University in Nijmegen, the University of Turku, the University of Namibia and the University of South Africa; and major international collaborations including the EHT and the Cherenkov Telescope Array Observatory (CTAO). Via the ERC project, the group is also involved in commissioning a new 14-metre radio telescope in Namibia, the Africa Millimetre Telescope.

The group’s research is built on a developing portfolio of in-house software: a GPU-accelerated general-relativistic magnetohydrodynamics (GRMHD) simulation code (in C, with GPU acceleration primarily on NVIDIA and AMD hardware); a GR ray-tracing code (in Julia) that produces images and spectra from those simulations, and is designed to conduct Bayesian parameter inference; semi-analytic jet and accretion modelling codes (in C++, with a Python interface) that are being integrated into the open-source Gammapy framework for use across the CTAO community and beyond; and increasingly building on other imaging and analysis frameworks used within the EHT Collaboration. These codes run on local, national and international high-performance computing (HPC) facilities.

This part-time 0.5 FTE Research Software Engineer (RSE) role will provide the professional software engineering capability at the centre of this network and will suit someone excited to work at the interface between modern software engineering and some of the most cutting-edge research in astrophysics.

The role-holder will establish and manage version-controlled repositories for all group software, deploy and optimise the codes on available HPC facilities, and embed good software practice across a distributed, international team through documentation, testing, code review and training. The role has a strong outward-facing dimension: coordinating shared code development with partner institutions and contributing the group’s codes to community open-source frameworks used by international collaborations.

The purpose of the role is to ensure that the group’s research software is reliable, performant, well-documented and reproducible; that development remains coordinated with partner groups as shared codes evolve; and that software engineering expertise is retained within the group as projects and personnel change. The role also offers clear scope for growth: the role-holder will be encouraged to engage directly in the group’s research, helping develop new computational methods and contributing to publications, with responsibilities that can expand as the Cambridge node grows.

• Establishing and managing reliable version-control infrastructure for the group’s codes, including migration of legacy repositories. Defining repository structure, branching, access and release workflows so that all group software is secure, documented, citable and reproducible.
• Liaising with partner development groups at other institutions to track external development branches of shared codes. Planning and carrying out the integration of selected external developments into the group’s versions, coordinating builds and testing.
• Setting up, testing and optimising the group’s numerical simulation codes on local, national and international (US/CAN/EU) HPC clusters using AMD and/or NVIDIA GPUs (e.g., via CUDA, HIP, OpenCL, or similar). Carrying out application profiling and benchmarking, and contributing technical cases and support to group proposals for acquiring more HPC computing time.
• Designing, coding, verifying, testing and documenting new physics capabilities within the group’s codes, together with the research team, profiling and optimising their performance.
• Investigating, prototyping and implementing machine-learning and advanced statistical methods (e.g. improved sampling for Bayesian parameter inference, surrogate models/emulators, simulation-based inference) to accelerate the group’s modelling and inference workflows; evaluating their performance and implements improvements.
• Supporting development of the group’s semi-analytic modelling codes (C++ with a Python interface) through performance optimisation and branch management, and planning and carrying out their integration into the open-source Gammapy framework, adhering to community coding standards, so that they can be adopted across the CTAO community and beyond.
• Implementing automated testing, continuous integration and documentation across the group’s codebases. Providing advice, guidance and training to researchers and students in both Cambridge and Amsterdam to promote adoption of software engineering best practice.
• Actively maintaining specialist knowledge and providing detailed advice to group leader Prof. Markoff on software strategy, computing resource planning and data-management plans for grant proposals and reports.
• Within the group’s research goals, building on and refining ideas for developing and evaluating new computational methods, collaborating with researchers on their application to open problems; reporting on work carried out and potentially contributing material of publication quality to group publications.

Salary: Grade 7

Full-time salary is £35,608-46,049 per annum and will be adjusted on a pro rata basis as this is a 0.5FTE position.

• In addition to the base salary, the successful candidate for this post will will receive an additional 2.5% supplement to their pay.
• This supplement will be paid until the conclusion of the Cambridge Pay Review Project.

The University of Cambridge is a signatory to the San-Francisco Declaration on Research Assessment (DORA), and in recruitment or promotion evaluations will assess research on the basis of its merits rather than the journal or venue in which it is published. Applicants should not include Journal Impact Factors or uncontextualized metrics in their applications. Full details are at: https://www.research-strategy.admin.cam.ac.uk/research-policy/DORA

The University actively supports equality, diversity and inclusion and encourages applications from all sections of society. The University of Cambridge thrives on the diversity of its staff and students.

Applications from underrepresented groups are particularly welcome. We have an active Equality and Diversity Committee which continually works to further the aims of the Athena SWAN charter.

The University has a number of family-friendly policies and initiatives, including a returning-carer scheme, childcare costs support, university workplace nurseries, university holiday play-schemes, and a shared parental-leave policy. The University has a responsibility to ensure that all employees are eligible to live and work in the UK.