Scientific Computing Officer — HPC Systems & Performance (ref. AD/26/09)

Job Summary

Required Qualifications

• •Basic programming or scripting skills in at least one language, such as Python, Bash, C, C++, Fortran or a similar language.
• •Basic knowledge of Linux and command-line environments.
• •General understanding of scientific computing or data-processing workflows.
• •Interest in computer systems, scientific software and application performance.
• •Ability to troubleshoot technical problems in a structured way.
• •Willingness to learn about HPC architectures, compilers, GPUs and parallel computing.
• •Ability to work as part of a technical team and communicate with researchers.
• •Working knowledge of English.

Desired Qualifications

• •Experience or knowledge in any of the following areas would be valuable, but is not mandatory: Computer architecture, including CPUs, memory hierarchies, caches, NUMA systems and storage I/O.
• •Experience or knowledge in any of the following areas would be valuable, but is not mandatory: Compilation of C, C++ or Fortran scientific software.
• •Experience or knowledge in any of the following areas would be valuable, but is not mandatory: Compilers such as GCC, LLVM, Intel oneAPI or NVIDIA HPC SDK.
• •Experience or knowledge in any of the following areas would be valuable, but is not mandatory: Selection of compiler flags based on CPU architecture and available instruction sets.
• •Experience or knowledge in any of the following areas would be valuable, but is not mandatory: Build systems such as CMake, Make, Autotools or Meson.
• •Experience or knowledge in any of the following areas would be valuable, but is not mandatory: Scientific software deployment using Spack, EasyBuild, Conda or environment modules.
• •Experience or knowledge in any of the following areas would be valuable, but is not mandatory: GPU computing using CUDA, ROCm, OpenCL or GPU-enabled scientific frameworks.
• •Experience or knowledge in any of the following areas would be valuable, but is not mandatory: Parallel programming with MPI, OpenMP or similar technologies.
• •Experience or knowledge in any of the following areas would be valuable, but is not mandatory: HPC workload managers, particularly Slurm.
• •Experience or knowledge in any of the following areas would be valuable, but is not mandatory: Performance analysis, profiling and benchmarking tools.
• •Experience or knowledge in any of the following areas would be valuable, but is not mandatory: Linux system administration, networking, storage or filesystems used in HPC environments.
• •Experience or knowledge in any of the following areas would be valuable, but is not mandatory: Containers such as Apptainer, Singularity or Docker.
• •Experience or knowledge in any of the following areas would be valuable, but is not mandatory: Scientific applications in areas such as molecular dynamics, structural biology, bioinformatics, imaging, artificial intelligence or data science.
• •Experience or knowledge in any of the following areas would be valuable, but is not mandatory: Automation, configuration management, continuous integration or infrastructure-as-code practices.
• •Experience or knowledge in any of the following areas would be valuable, but is not mandatory: Git and collaborative software-development workflows.