This is the multi-page printable view of this section.
Click here to print.
Return to the regular view of this page.
CompSys 2025 (May 2025)
Welcome to CompSys 2025, a Computer Science conference designed to highlight Dutch Computer Systems and Networks research,
while fostering and strengthening national and international collaboration.
We aim to provide a meeting space for network, computing, and computer system research and industry ideas.
Latest news and announcements:
- Thank you all for attending CompSys 2025! We had a great edition, as you can see from the many nice pictures . See you next year!
- The CompSys2025 best presentation award was awarded to Efficient Hyperparameter Tuning for Auto-tuning (Floris-Jan Willemsen, Rob van Nieuwpoort and Ben van Werkhoven). Congratulations!
- The slides from our who-is-who session are available here!
- You can vote for the Best presentation award of CompSys2025 using this form!
- Welcome to CompSys2025! Our welcome presentation can be found here!
- We are organizing a "who is who" session! If you are interested to give a 2-minutes pitch about your research interests, please fill this form by Monday May 19 (noon)
- The CompSys2025 program is online -- welcome!
The 8th edition will take place on 21-23 May 2025, in collaboration with
NCCV, the Netherlands Conference on Computer Vision under the auspices of the ASCI research school.
The location for the 2025 edition of CompSys is the Carlton President Hotel in Utrecht (Floraweg 25, 3542DX Utrecht).
Building up on the success of the previous seven editions (2024, 2023, 2022, 2021, 2019, 2018, 2017), the 8th edition of the
conference will emphasize efforts on community building and providing a forum to discuss ongoing and future projects
among all members of academic research groups in the Netherlands.
The conference will focus on the major research and practice themes related to computer systems. We envision a
diverse program, featuring keynotes on advanced topics, strong scientific contributions, and
exciting early ideas. We strive for diverse participation from all the interesting and interested parties in the Netherlands,
and we especially welcome senior members of the research community, junior faculty members, PhD, master, or undergraduate students.
2 - Keynotes
We have the pleasure of welcoming our keynote speakers at CompSys'25.
| Wednesday May 21, 10:00 |
| Title:
IT & environment: an inconvenient combination?
|
| By Bernard van Gastel (Radboud University) |
Abstract:
We are digitalizing our whole society, from elderly care to our school systems. We are connected constantly, and there seems to be no limit to our need for digitalization and computation. But there are side effects. What are the effects on our environment?
This presentation will give an overview of the resource consumption of IT in the Netherlands, both in energy and water. It will show a number of trends, both in energy consumption and approaches to minimize our environmental impact. It closes with naming a number of challenges for computing scientists.
|
| Short bio:
Bernard van Gastel is an assistant professor Sustainable Digitalization working at the computing science institute of the Radboud University. Besides his research in analysis methods to predict the energy consumption of software and make that info available to software developers, he researches sustainability in a broader sense. This includes the effects of software on society, such as privacy. He often works together with academics from other disciplines.
Presently, he is quartermaker for sustainability at the Institute for computing and information sciences (iCIS), for both research and education. There he set up the Software Energy Lab to measure the energy consumption of software. On the education side, he is integrating sustainability in the curriculum of computing science.
Regularly his work results in societal impact. For example, he found errors in solid state drives of vendors making two thirds of the world wide storage devices. This resulted in adjustments in the Windows operating system and storage devices itself. He designed a large system for privacy friendly storage of medical research data, so that the impact of data breaches is minimized.
|
| Thursday May 22, 9:00 |
| Title:
Towards Designing Systems for Uncompromized Efficiency
|
| By Jan Rellermeyer (Leibniz University Hannover) |
| Abstract:
The exponential performance improvements in the era of Moore’s Law and the commoditization of computer resources through cloud computing have led to liberal spending of energy on computation, especially big data processing. The availability of a developer workforce willing to deal with the ever-increasing complexity of systems has so far concealed the growing divide between hardware and software. I term this trend the Efficiency Gap and believe that it is going to significantly shift future priorities in computer science. It is time to re-visit the entire system stack and prepare for the new era in which efficiency matters--beyond niches like High-Performance Computing. In this talk, I will outline my vision for researching new design principles to structure large-scale data processing systems in a way that enables higher efficiency without burdening developers. I challenge multiple assumptions of contemporary system stacks like the treatment of applications as black boxes and the orthogonal scheduling of resources. Many of them are based on fundamental abstractions that have been largely inherited from the 1980s but now inhibit efficient use of modern and emerging hardware. As an alternative, I propose a new design principle called Exo-Modularity which exposes critical application structure to the system and enables system software to make smarter decisions on resource allocation and scheduling. As a result, the system is put into a position to automatically optimize the resource allocation by prioritizing parts of the application where the flow of data is bottlenecked, thereby unlocking the true potential of the hardware and closing the efficiency gap.
|
| Short bio:
Jan S. Rellermeyer graduated from the Systems Group at ETH in 2011 under the supervision of Gustavo Alonso and Timothy Roscoe. Afterwards, he worked for IBM Research in Austin, TX as a research scientist for several years and served as an adjunct member of the faculty at The University of Texas at Austin. In 2017, he returned full-time to academia by joining TU Delft as an Assistant Professor in Distributed Systems. Since 2022, Jan is now Full Professor and Chair for Scalable and Dependable Software Systems at Leibniz University Hannover. He has contributed to different open-source projects under the Apache and Eclipse Foundation and was project lead of the Eclipse Concierge project. Furthermore, he was active in standardization bodies like the OSGi Alliance in which he held the position of an invited researcher from 2008-2011. Together with his co-authors, Jan received Test of Time Awards of the 2017 ACM/IFIP International Middleware Conference for just work on the R-OSGi middleware and of the 2022 IEEE/IFIP International Conference on Dependable Systems and Networks (DSN) for his work on studying and mitigating robustness issues in the Android inter-component communication system.
|
| Thursday May 22, 14:00 |
| Title:
The Era of Compilers and DSLs
|
| By Stephan Herhut (NVIDIA) |
| Abstract:
The physical limitations of chip manufacturing have led to increasingly heterogeneous and distributed architectures. While these designs have enabled significant advancements in performance and computational capabilities, we can no longer automatically benefit from such progress solely through improvements in architecture design. Instead, software must be specifically tailored to new architectures to leverage their capabilities, necessitating the co-design of software and hardware. Efficiently mapping algorithms to architectures now requires specialized tools, either to automate the process or to enable engineers and researchers to directly tailor algorithms. Consequently, there has been a proliferation of compilers and domain-specific languages (DSLs). In this keynote, I will review the evolution of hardware and discuss why no single tool or framework can fully enable programmability. Embracing the plurality of tools is inevitable, and our goal should be to build a cohesive platform that facilitates reuse for users and supports the development of systems software.
|
| Short bio:
Stephan Herhut is an accomplished compiler engineer and currently serves as Distinguished Compiler Engineer at NVIDIA, driving compiler architecture. His experience includes nearly ten years at Google, where his last role was Staff Software Engineer, leading the development of XLA CPU and GPU machine learning compilers. At Google, he also developed compilers for platforms and languages such as V8, Android, Dart, MLIR, and embedded systems. Earlier in his career, his research at Intel Labs explored auto-parallelization techniques and language design, with a focus on functional programming languages. He is passionate about sharing his knowledge and insights into compiler design.
|
| Friday May 23, 9:00 |
| Title:
Integrating Resilience and Sustainability into Mobile Networks
|
| By Suzan Bayhan (University of Twente) |
| Abstract:
While mobile networks evolve to support increasing demands for capacity, low latency, and diverse applications, traditional design goals focused solely on efficiency and performance are no longer sufficient. Two critical challenges drive the need for a broader design perspective. First, the climate crisis calls for design and operation of mobile networks with a sustainability objective. Second, the growing reliance on mobile connectivity from operation of critical services to daily activities demands enhanced network resilience. In this talk, I will discuss how the principles of sustainability and resilience can be embedded into mobile network architecture and operation. In particular, I will highlight strategies such as multi-connectivity and infrastructure sharing that can help achieve these objectives.
|
| Short bio:
Suzan Bayhan is an Associate Professor at the University of Twente, the Netherlands. Her research interests include mobile networks and sustainable communication/computing continuum. Before joining UT, she was a senior researcher at TU Berlin, and a postdoctoral researcher at the University of Helsinki, where she is currently a docent in computer science. With co-authors, her research papers received a best paper award from ACM ICN and IEEE WoWMoM, and a best demo award from IEEE INFOCOM. Suzan has served as the N2Women mentoring co-chair in the past and currently she is the Scientific Director of 4TU.NIRICT.
|
3 - Panel Discussion
How to Valorise sustainability in Computational Systems?
Sustainability is a term that we increasingly face everywhere: from sustainability goals to sustainability regulations and sustainability investments. Given the increasing share of computer systems in the overall societal power demand, there is clearly a need to address sustainability aspects from a CompSys perspective.
The key question that we try to address in this panel is: What are we currently doing? What can we do better / more of? What do businesses need? What does the society need? To what degree do these align?
How can we quantify / predict the possible impact of such measures? In particular when it comes to aiming for funding, one of the key challenges is to attribute concrete values to sustainability goals.
We try to bring together stakeholders from industry, political executive, and academia to share their view and discuss these matters with the CompSys community.
The key panelists will be:
- Jason Maassen (Netherlands eScience center) - Jason is Technology Lead for Efficient Computing at the NL eScience center. Jason is involved in many of the projects at the Netherlands eScience Center that apply parallel and distributed programming to scientific applications. In addition, he guides internal software development at the eScience Center and scouts for new software technology that can be used in new projects.
- Sagar Dolas (SURF) - Sagar Dolas is a program manager at SURF, where he leads and manages various programs and initiatives focused on Future Infrastructure and Applications, Advanced Computing, and Networking. In recent years, he has concentrated on energy as a design principle for the future of computing. He has led projects to integrate energy management tools with the Dutch supercomputer Snellius.
- Bernard v. Gastel (Radboud University) - Bernard van Gastel is an assistant professor in Sustainable Digitalization working at the computing science institute of the Radboud University. Besides his research in analysis methods to predict the energy consumption of software and make that info available to software developers, he researches sustainability in a broader sense. This includes the effects of software on society, such as privacy. He often works together with academics from other disciplines.
- Rachel Kuijlenburg (Logius, Dutch government IT Department) - Rachel Kuijlenburg is the sustainability coordinator at Logius. She is tasked with developing a sustainable procurement strategy for the IT products being bought by Logius, which involves creating alignment over many internal departments. Two procurements have included explicit conditions for environmental sustainability now, which will also be incorporated into new procurement contracts. Before she moved to Logius, she was a teacher at the Haagse Hogeschool and a researcher for the Mission Zero center, specializing in circulair business management.
The moderator of the panel will be Sven-Bodo Scholz (RU)
5 - Town hall
To be announced. Stay tuned.
7 - Location
Location:
The location for the 2025 edition of CompSys is the Carlton President hotel in Utrecht.
8 - Submission details
Submission Guidelines
All contributions will be reviewed by the Program Committee. Accepted
contributions will appear in the final program either as a short talk
or a full presentation, depending on the reviews. All presentations
will be made available in digital format, unless otherwise instructed
by the authors. For submission, the PDF format is mandatory.
To foster the broadest possible engagement and exchanging of ideas,
CompSys 2025 does not claim copyright, making it possible for authors
of accepted contributions to present work that has already been
published or is in the process of being published elsewhere.
Important dates
| Description |
Date |
| Paper submission deadline |
8 April 2025 (EXTENDED!) |
| Author notification |
25 April 2025 |
| Registration deadline |
10 May 2025 |
| In-person conference |
21-23 May 2025 |
Submission types
CompSys 2025 welcomes three types of contributions: research papers, work-in-progress papers/early ideas, and negative/failed research results.
Long papers
Research papers on your best results from the past year(s). This includes papers already submitted to and/or accepted at (inter)national conferences or workshops (please indicate the original venue in the submission form). Long papers must not exceed 12 pages in double-column format or 15 pages in LNCS format and can be submitted using any of the commonly used templates (e.g., ACM, IEEE, LNCS).
Short papers: Work-in-progress and early ideas
Since CompSys is a forum that encourages discussions about early and exciting ideas, we specifically welcome extended abstracts highlighting early ideas and work-in-progress papers. These submissions are particularly suitable for graduate and undergraduate students working on their theses or PhD students who have recently started and want to share preliminary results with the community. In particular, we encourage contributions in the form of short talks to share early, unexplored ideas with the community, stimulate discussions, and gather feedback. These talks may be especially interesting for early-stage researchers. Submissions require a short paper of at most 2 pages (not including references) in IEEE double-column format or 4 pages (not including references) in LNCS single-column format. The paper should mention the research question being addressed, outline the novelty and/or originality of the idea or approach, and contain a summary of preliminary results.
Negative research results
As in previous years, we also solicit contributions sharing negative results, incorrect methodologies, or invalidated hypotheses to share lessons learned and remind ourselves that trying ideas that do not lead to expected outcomes is a normal part of research. Submissions of negative research results papers require a short paper of at most 2 pages (not including references) in IEEE double-column format or 4 pages (not including references) in LNCS single-column format.
Submission Portal
9 - Contact
International Conference on Computing Systems
Paula Diks (ASCI OFFICE)
Van Mourik Broekmanweg 6
2628 XE Delft
Netherlands
