Interview with Chris Broekema, senior researcher in high-performance computing at ASTRON

First ODISSEE hackathon

Exploring sustainable computing for next-generation radio astronomy

Could you briefly introduce yourself and ASTRON?

My name is Chris Broekema. I am a researcher in high-performance computing at ASTRON, the Netherlands Institute for Radio Astronomy. Our mission is to make discoveries in radio astronomy happen. Here in Dwingeloo, in the Netherlands, we have an institute of around 200 people with the capability to design, build and operate a radio telescope from end to end. We have antenna engineers, analogue electronics engineers, digital engineers, computer scientists, software engineers and astronomers working together in-house.

« A major objective of ODISSEE is sustainability and energy efficiency. »

You have worked on LOFAR and SKA infrastructures. What was your role there?

For LOFAR, my personal work focused mostly on the design of the high-performance computing infrastructure. In the early 2000s, I worked on the network infrastructure and central compute infrastructure, including the digital correlator and beamformer. The first implementations were based on IBM technology, and later generations moved towards GPU-based infrastructures.

I also worked on the Square Kilometre Array, where I was responsible during the pre-construction phase for the hardware design concept of the Science Data Processor.

Turning to ODISSEE, what contributions does ASTRON bring to the project?

A major objective of ODISSEE is sustainability and energy efficiency. At ASTRON, we have a unique ability to measure the energy consumption of computational components at very high resolution. Standard tools usually measure energy consumption at the level of seconds, whereas we can go down to the microsecond level. This allows us, for example, to identify where different computational kernels are launched simply by looking at the energy profile of a GPU.

In addition, we contribute to the characterisation and evaluation of novel computing architectures. This effort is closely tied to one of the main objectives of the ODISSEE project: fostering collaboration and hands-on experimentation across communities, which is why this hackathon played a central role. It provided a concrete setting to test emerging technologies together and explore their potential in real scientific workflows. We are particularly interested in the NextSilicon Maverick-2 device, and was a key focus of experimentation during the hackathon, and in the SiPearl Rhea processor, that will hopefully be available this year.

« At ASTRON, we have a unique ability to measure the energy consumption of computational components at very high resolution. »

And what did you learn from working with the NextSilicon Maverick-2 device during this ODISSEE hackathon?

The technology is quite unique and promising. By characterising this hardware with demanding scientific workloads, we can provide valuable feedback to the vendors. They are interested in understanding how we want to use their equipment, which can help guide future developments.

This is also important from a European perspective. We want to reduce reliance on single suppliers for computational resources, and technologies such as NextSilicon and SiPearl are promising in that respect.

« The hackathon created an opportunity for people from CERN, and SKAO-related communities to discuss what could be achieved together. »

The ODISSEE project aims to build long-term synergies between high-energy physics and radio astronomy. How does this hackathon contribute to that vision?

High-energy physics and radio astronomy are very different scientific domains, but both are extremely computationally intensive. They solve different problems, but some of the computational primitives they use are similar. With enough effort, it may be possible to reuse methods, tools or even code across domains.

The hackathon was part of this broader vision. It created an opportunity for people from CERN, SKAO-related communities and other partners to meet in person and discuss what could be achieved together. The first objective was to increase collaboration. There was a need to strengthen interaction between institutes and scientific communities. It has also been very interesting to gain more insight into the challenges faced by the high-energy physics community. We have also been fortunate to work with Simula, the Norwegian computer science research institute, which has been very effective in hosting experimental machines and providing access to its clusters.

How do you approach the relation between software and hardware?

We try to identify the most suitable hardware for a given workflow and adapt the workflow to that hardware. Code optimisation is generally hardware-specific: software has to be tailored to each architecture.

At the same time, we also try to work in the other direction. In traditional hardware-software co-design, software is often modified to fit hardware that has already been selected. In ODISSEE, and in potential follow-up systems, we aim to make this process more bidirectional. This means not only adapting software to the hardware, but also considering how hardware components could be selected or configured to better match the performance characteristics of the software. The goal is to achieve a more efficient combination of hardware and software.

You mainly focus on energy efficiency in the project. What drives your approach to designing more efficient infrastructures?

For me, the main goal is to reduce the environmental impact of scientific discovery. In radio astronomy, scientific discovery is increasingly limited by the amount of computing we can afford, both financially and environmentally. Producing a scientific image from raw data can require a significant amount of cluster time, and this cost may become prohibitive in the near future.

One of our objectives is therefore to make the cost of scientific discoveries visible, to demonstrate that cost, and then to reduce it. The aim is to allow astronomers to make more discoveries within the same energy budget.

Projects like ODISSEE bring together different scientific and technical communities, allow us to test emerging technologies on real workloads, and help us think about how to make future scientific computing more efficient and sustainable.

« The aim is to allow astronomers to make more discoveries within the same energy budget. »

News and events

Follow us

View More
This project has received funding from the European Union’s Horizon Europe research and innovation program under grant agreement N°101188332. This website reflects only the author's view and the Commission is not responsible for any use that may be made of the information it contains.

Useful resources

Privacy & Cookie
The Project
The Pilot Program
Webinars
Deliverables