Interview with Bernd Mohr
Connecting Europe’s post-Exascale strategy with the world
Europe cannot develop a post-Exascale strategy based solely on its own interests. The United States is accelerating around AI, China and Japan are investing in sovereign hardware and large-scale infrastructures, and emerging regions are developing creative approaches with fewer resources. For Bernd Mohr, Division Head “Application Support” at Jülich Supercomputing Centre, this is precisely why international collaboration matters: Europe needs to understand where others are going to decide where it can lead, where it should invest, and where cooperation is essential.
As part of the Strategy for Post-Exascale (SPE) project, Forschungszentrum Jülich is leading WP3 – International Collaborations, which aims to compare Europe’s strategy with global developments and strengthen Europe’s role in international post-Exascale discussions. Bernd’s extensive experience in HPC and international collaboration is invaluable to this work.
What is your background, and what does Jülich bring to SPE?
I have been working in high-performance computing since the mid-1980s, even before it was called HPC. My main research area is performance tools: software that helps analyse how applications run on supercomputers, where they waste time, where communication is inefficient, and how users can improve their codes. At Jülich, I became involved in many international collaborations. I worked with major computing centres in the US, China, Japan, Latin America and elsewhere. This network is one of the reasons I am involved in SPE.
Forschungszentrum Jülich is also one of Europe’s major HPC centres. It is part of a large national research facility, with around 7,500 people working across physics, chemistry, material science, medicine and many other fields. The computing centre supports real scientific applications and real-world problems. Today, Jülich hosts JUPITER, Europe’s first exascale-class system.
« Europe is very strong in software, and this is one of its assets. »
What is changing internationally in HPC, AI and post-Exascale computing?
At a high level, all major regions recognise the same trend: current scientific applications, such as weather, climate and many others, require huge computing capacities. This need becomes even stronger with AI. AI data centres from major companies are sometimes even larger than the systems we operate.
But each region has different priorities. In the United States, the discussion is now strongly dominated by AI, because of companies such as Google, OpenAI, Meta and others. HPC is still there, but AI drives much of the agenda. Japan and China are closer to Europe in the sense that they have strong government programmes, large centres and major machines. Like Europe, they are also working on sovereignty: developing their own hardware, processors and models to become more independent from US dominance.
Europe has its own challenge. We have many countries, many opinions and many priorities. This makes decision-making harder. But it also gives Europe more openness and discussion. Europe is also very strong in software, and this is one of its assets.
Why is international collaboration necessary for SPE?
The main purpose of SPE is to come up with a European plan, European priorities and a way forward. To do this properly, we need to know what others are doing. This helps us identify our own strengths. We should not invest large amounts of money in areas where we cannot win, unless they are strategically important. For me, this is the role of WP3: to monitor what other regions are doing, understand international priorities, and bring this knowledge back into the European discussion.
In return, it also positions Europe as a leading actor in the global landscape and helps to push European priorities within the frame of a frank and open dialogue. We contribute to make Europe a valuable partner, which is one of our clear asset in today’s international relations.
We do this by participating in international meetings and sometimes organising them. A key example is InPEx, the International post-Exascale initiative, which originally connected Europe, the United States and Japan, and now also includes Latin American partners. Next year, the major InPEx meeting will take place in Europe, in Munich, in April. This will be an important moment to bring global perspectives into the post-Exascale discussion.
« The main purpose of SPE is to come up with a European plan. To do this properly, we need to know what others are doing.»
What is the main challenge in bringing HPC, AI and quantum together?
The computing landscape is becoming more complicated. We now have to combine traditional large-scale computing, emerging quantum technologies and the rapidly changing influence of AI. This is both frustrating and exciting. We discuss ideas, and by the time we are ready to write a paper or proposal, the AI community arrives with another new concept. Agentic AI, for example, was a small dot on the radar a year ago.
The real challenge is to bring these three areas under one umbrella. Traditional HPC will still be needed: we still have to design airplanes, run climate simulations and solve major scientific problems. AI can accelerate some of these methods and create new combinations. In the longer term, quantum technologies may also help accelerate specific tasks.
How can Europe remain open while strengthening its own position?
Europe cannot define its strategy without watching what others are doing and talking to them. There are open-minded people around the world, including in the US and China, and within the limits of what is possible, we can still exchange ideas.
But AI makes this more complex, because it has a clear impact on industry and the economy. Europe needs to remain open, but not too open. We need to find the right balance between learning from others and helping Europe excel.
No region has everything it needs to lead everywhere. Different regions have different strengths. The challenge lies in identifying areas in which Europe can take the lead, areas in which we should collaborate with or learn from others, and ways in which international collaboration can benefit everyone.
« The challenge lies in identifying areas in which Europe can take the lead, and areas in which we should collaborate with. »
Fill in the form
Join Us to Help Shape the Future of Computing
SPE project
This project has received funding from the European Union’s Horizon Europe research and innovation program under grant agreement N°101296624. 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.