GPU consultant/scientific programmer

GPU consultant/scientific programmer

Published Deadline Location
28 Apr 12 May Amsterdam

You cannot apply for this job anymore.

Browse the current job offers or choose an item in the top navigation above.

Job description

Do you want to work in a challenging and innovative environment? Would you like to contribute to new developments in artificial intelligence, high performance and specialised computing? If so, we welcome you to join SURF, the collaborative organisation for ICT in the Netherlands.
We are currently looking for a consultant who can develop GPU specific code, participate in (inter)national projects and support our user community from the High Performance Machine Learning team.

What does the position entail?
The High Performance Machine Learning team participates in a number of national and international projects in the field of artificial intelligence and offers support and expertise to members and partners. Using a two stepped approach, we accumulate knowledge in order to disseminate and aid our community. Our involvement in various consortia consists of enabling academia, public, and private parties to benefit from computing and data infrastructures, as well as researching and co-developing new data driven methods for large scale analytics. We are actively disseminating our acquired expertise to the scientific community. This usually involves publications, participating in courses, webinars, and public talks.
As a GPU consultant/scientific programmer at SURF, you will be part of the department’s operational team. You will support the development of efficient High Performance Computing workflows on GPUs, i.e. performance optimization and scaling. You will be part of a young and multicultural team collaborating with the research community to develop a set of services around artificial intelligence solutions.

What will you be doing exactly?

  • Interact with academic, public and private stakeholders in order to support research, as well as co-develop and benchmark software on large scale and specialised compute and data infrastructures;
  • Participate in projects at the intersection of high performance computing, cloud computing, specialised computing and various scientific fields, e.g. medicine, chemistry, systems;
  • Participate in the outreach activities that could involve developing course materials, teaching, hosting webinars, organizing hackathons, disseminating scientific and engineering results, releasing public code, etc;
  • Keep up-to-date with relevant information in the fields of HPC and GPU programming.

A few examples of previous projects:

  • Application of transformer models for chemical reaction predictions;
  • Semantic segmentation for histopathology on gigapixel images;
  • Performance analysis of Nvidia GPUs vs AMD GPUs for ML;
  • Deducing the properties of galaxy mergers using variational auto-encoders;
  • Using GANs to improve the quality of CT scans.




Your skills include the ability to apply your in-depth knowledge of specialised computing and your hands-on experience for practical problem solving. You enjoy working in an independent manner and have the ability to work in a team. You have excellent interpersonal and communicative skills and a good spoken and written command of English and Dutch. You’re analytically strong, proactive and recognize yourself in the following:

  • An academic level of thinking and working: (MSc/PhD level) in computer science, mathematics or other related fields. This can be equated to professional experience;
  • Scientific and/or practical background in developing (GP)GPU codes, involving CUDA and/or OpenCL and/or PGAS languages codes. Other architectures are considered a plus;
  • Strong interest in the application of specialised computing to solve real problems;
  • Knowledge about parallel computer architectures and programming languages e.g. MPI;
  • Programming experience (at least C/C++/Java and/or Python);
  • UNIX/Linux knowledge;
  • Knowledge of scripting languages (e.g. shell scripting, Perl).

Conditions of employment

Fixed-term contract: One-year.

  • Educational background: academic
  • Location: Amsterdam/virtual
  • Employment: 32-40 hours
  • Closing date: 12 May 2021

What we offer
A diversified and challenging job for 32-40 hours (0.8 - 1.0 FTE) in an informal and collegial atmosphere with a high ambition level. SURF offers extensive training opportunities and excellent fringe benefits, including 36 holidays and a 1st class NS business card. This position has a salary range of € 3.535, - to € 5.802, - gross based on a full-time appointment. Our office is located at the Science park Amsterdam, with good public transport facilities nearby. The appointment is initially for a one-year contract.



SURF supports research and education through innovation and service development, by leveraging its expertise in High Performance Computing, Data Processing and Data Services.


Research Services

Within SURF, the department Research Services is responsible for the national supercomputer Cartesius and its successor, Snellius, as well as the national compute cluster, Lisa. Cartesius currently has 50,000 CPU cores and 132 GPUs, with a total peak performance of 1.8 Pflop/s. Lisa currently consists of almost 8,000 CPU cores and 104 GPUs. Snellius will be the new Tier-1 Dutch National Supercomputer with a peak performance of 14 PFLOPS/s, to become operational in 2021. The majority of our users are researchers at Dutch Universities, members, as well as international research institutes.


  • Research, development, innovation; Education; Technical and laboratory; IT
  • Natural sciences; Engineering
  • 32—40 hours per week
  • €3535—€5802 per month
  • University graduate
  • SP GPU consultant


Science Park 140, 1098 XG, Amsterdam

View on Google Maps

Interesting for you