The HPC Council

Computing is of increasing importance to the success, impact and productivity of research. For many disciplines, computing is either a sheer necessity, or an exciting opportunity to open up new avenues of research.

The VU has installed a High Performance Computing (HPC) council in 2015. This council gives solicited and unsolicited advice to the VUO regarding the strategic direction of research computing services the VU offers. Secondly, the HPC Council distributes annual grants (max 10k Euro) via a call for proposals.

We invite VU researchers to come up with innovative ideas on how to use research computing to improve scientific workflows in your domain.

The importance of high-performance computing is beyond dispute. By using high-performance computers and computational methods, data can be processed at lightning speed, which in turn can speed up research. Within the VU, we use services such as BAZIS, Lisa, and Snellius for this purpose, but researchers also make use of relatively 'new' services, such as commercial cloud parties. Development costs or time are a barrier. With this call, we want to get more ideas from researchers to the surface and subsidize them so they can be implemented.

Budget and criteria
Do you have an idea of how to speed up or improve research with the help of computational methods? If so, send your proposal to the HPC Council before 1 December 2023. First check if your proposal is eligible for financial support and meets the following criteria:

The project leads to an improvement in the use of computational methods; whether in your research domain, within your department, or for a limited number of HPC users.
The developed product is sustainable (for example by using widely accepted standards, but also by a good organization of support and maintenance).
The costs of your project are clearly defined. The maximum amount that can be applied for is 10,000 euros. Applications can be submitted for hardware, personnel, and cloud costs.
How you present your idea/project is up to you. The VU community must understand its importance.

Applications should meet these four criteria and should not exceed three A4 pages. You can use this template. You can send your application to the secretary of the HPC council: Erik Koedam. Erik is also available for additional questions.

Meetings
The HPC Council has 4 annual meetings. The dates will be listed here annually to help you determine when to submit your call for proposal.

16-03-2023
15-05-2023
31-08-2023
23-11-2023

Additional information

A few ongoing developments will impact the requirements on research computing in the coming years. These developments are closely monitored by the HPC Council and are taken into account for the optimalisation of the research computing services the VU offers.

Increasingly broad adoption of computing
More research disciplines are adopting research methodologies requiring extensive analysis and computation. The increasing adoption of topics such as machine learning, data analysis, natural language processing and image processing will pose increasing demands – in scale and performance – on research computing solutions.
Wave of adoption of machine learning techniques
Researchers almost universally brought up the increasing enthusiasm for machine learning (especially deep learning) techniques in research. The characteristics of machine learning techniques – lots of usually simple calculations, often with large datasets as input – means that demand for GPU-based computing capacity and accompanying data storage is likely to significantly increase over the coming years.

Additionally, many of the researchers newly adopting these techniques will come from traditionally non-technical backgrounds, with little experience working with Linux and shell scripts.
Increase in data available to researchers
With researchers’ having access to increasingly large amounts of data, access to computing capacity is becoming an increasingly important capability for researchers.

Relevant links:

Research computing
Here you can find specific links and / or request forms that are specifically relevant to you as a researcher.
Snellius
The Dutch National Supercomputer Snellius is the flagship of SURF's research services. It is the most comprehensive system in the field of capability computing in the Netherlands. Snellius is especially in high demand for its combination of fast processors, GPUs, and internal network, large storage capacity, and the ability to process large datasets.

Members HPC

The council aims (1) to have representatives of all fields of use of research computing, (2) to have representatives of the faculties that apply research computing and (3) to have a representative of the UvA in the VU HPC Council and vice versa.

Peter Colijn (chairman)

Head of Infrastructure Services, Department IT, Vrije Universiteit Amsterdam

Luuk Visscher (member)

Theoretical Chemistry, Faculty of Science, Vrije Universiteit Amsterdam

Dorret Boomsma (member)

Biological psychology, Faculty of Behavioural and Movement Sciences, Vrije Universiteit Amsterdam

Danielle Posthuma (member)

Complex Trait Genetics, Faculty of Behavioural and Movement Sciences, Vrije Universiteit Amsterdam

Charles Bos (member)

Finance, School of Business & Economics, Vrije Universiteit Amsterdam

Henri Bal (member)

Computer Systems, Faculty of Science, Vrije Universiteit Amsterdam

Boy Menist (member)

Head of ICT, Faculty of Science, University of Amsterdam

Jef Trimbos (member)

IT for Research, Department IT, Vrije Universiteit Amsterdam

Erik Koedam (secretary)

IT for Research, Department IT, Vrije Universiteit Amsterdam

e.koedam@vu.nl