HPC GPU Programmer

At DUG, we offer high-performance computing as a service for a wide variety of clients. Built on many years of experience in geophysics, our world-class compute cluster is tailor-made for scientific computing and attracts a diverse range of computationally-intensive projects.

We’re looking for a GPU programmer with familiarity in writing and maintaining optimised software for a variety of platforms, to help customers bring their GPU workloads to our HPC infrastructure.

As HPC GPU Programmer, you’ll be part of a team of technical experts responsible for ensuring that customers can use our compute infrastructure as productively as possible. Your pragmatic approach will help you deal with third-party software that's not always of especially high quality. You’ll become proficient in the languages and techniques used in these projects, as well as an evolving and somewhat-unique HPC environment.

Your tasks

• porting to and ongoing maintenance of scientific applications on DUG’s HPC environment
• inspection and maintenance of software written by other members of the team
• achieving a basic understanding of our customers' scientific domains
• acting as 3rd-level technical support, for otherwise-unresolvable customer issues
• providing and receiving regular, constructive feedback to and from your peers
• collaborating on priority and scheduling for rapid project delivery

Required

• strong C/C++ skills, along with at least one GPU-targeted toolchain (ideally CUDA)
• experience using GPGPUs for numerical processing in a commercial or academic setting
• excellent written and spoken business and technical English, suitable for working directly with clients when necessary
• impeccable professional references

The most successful candidates will also possess some or all of the following

• significant expertise as a Linux user/administrator (able to perform tasks such as installing complex software with challenging dependencies; building and installing a patched kernel; or diagnosing a device driver problem)
• experience optimising for both NVIDIA and AMD GPUs, and x86 CPUs
• a background or interest in some branch of science
• sufficient maths to be able to implement algorithms described in scientific papers