Following a keynote presentation at HiPEAC 2025, Prof. Giovanni de Micheli was interviewed about the future of computing from a European perspective.
Do you feel developments in Europe are on par with what is happening around the world in Computing?
I think Europe can do better and I think it must do better because everything that we do today, from traffic to health, goes through some sort of computing and we need to be able to own more of the technology. We have great companies we have great silicon foundries but I think that we can do more.
To do more also requires new thinking about computer architectures. In your keynote speech this morning you mentioned the Phoenix of computer architecture...
The notion of the Phoenix is that computing reinvents itself. Computing can be reborn out of its own ashes and that is important because we will have new technologies, we will have new ways of making computing tomorrow and we have to make that happen in these times so we have to be able to use the computing infrastructure that we have today. We have to use the talents that we have today in Europe in order to do new emerging design.
This is all about Europe but you have a long career, you've worked at Stanford. We're talking about students, we're talking about new people, new ideas but what in your career has been vital in understanding what we need? Can you recall a light bulb moment in your whole career or in your research?
It is about people I have met, their enthusiasm that drives one to work, with passion.
I remember at Stanford, Robert Dutton and John Hennessy, who hired me, and Patrick Abeischer who hired me at EPFL are all people who are ready for the discussion, who are ready for new ideas, who are ready to throw up what they've been planning to do for many years if you have something that can be new and better equipped to solve the problem.
So in many ways it is the passion but also it is the research and the network of people you're working with. That is also I think part of these high-level conferences. You're meeting up with people and getting new ideas. It's an environment where you need to get to a critical mass with people, discussing problems, and when things start rolling it goes by itself, you throw an idea to a graduate student and he or she will come up with another idea, and another one will pick it up from there and I tend to think that some of the best graduate students are people who didn't do what I told them to do.
And that is because they discovered something completely different, a different angle to what you've been telling them?
Well I have two examples. We were working on devices and we discovered interesting properties in some majority gates and one of my students, Luca Amaru, had this brilliant idea of reformalizing the majority algebra and making a tool that would work in CMOS, as well. That was not part of the plan, but it was extremely impactful on Academia and Industry, because Industry is using this tool.
The other person who has also been diverting away from my plans is Jerry Yang. He and David Filo were working in my group on Synthesis and they had an idea. One day they came and told me "Professor, we are making this company: we're going to list the various things that you can find on the Internet." I told him, "Well, we can list patents, inventions, technical papers..." and he looked at me and told me "No, no we are going to list cleaners, restaurants, car dealers, things that everybody would need on a daily basis."
And I told him "Well that's probably very trivial, it's like the Yellow Pages," and he told me: "Exactly. I'm going to put the Yellow Pages online." So they founded Yahoo! And the rest is history.
So, you know, it all comes down to research and students.
If you have, say, the European Commission and you're looking about at what is happening in Europe, where would you put your money, in terms of fundamental research: in Quantum Computing, in students, in the total system of companies working together? What do you feel is important in creating a breakthrough in Europe?
If you read the Draghi report on European technology, there are two things that come across strongly. Funding technology relating to chip design, including the manufacture of computing systems; and creating an environment where people are hiring the talent. We are still low on talent in Europe. We need to do this and to create enthusiasm. We need newspapers to talk about it, to bring people the notion that we are doing wonderful things in Europe, like cars and planes, with technology done here. We need to make sure that students stay in engineering and computer science, and apply their skills to making a better Europe.
I would like to take you back to the the road map of of computer architecture where we have, in history, x86 architecture and traditional computing. What are the technologies you are looking at now to bridge this gap? What is important, where should we look?
I think that the important point is heterogeneity and diversity, because our systems are very far away from the old computers that were actually solving one type of problem. Most of the computers that we use are embedded computers that work in connection with all kinds of electronic or mechanical devices, and there is the fact that specific functions can best be implemented in a special technology.
It's extremely important again to think of the Automotive Industry, where you need to have control which is plastic, and then you need to have signals that let high currents through, so that you need to have mixtures of technology like silicon and silicon carbide, for example, or silicon and gallium nitride, and it's extremely important that this happens in Europe: it is still not well-known enough.
Then in the distant future, we're looking at Quantum Computing. I heard you saying you know Quantum Computing is not for everyone. What will be the typical field where you will first recognize what Quantum Computing can offer?
I think that the first practical application will be related to physics and chemistry, to be able to understand how to control or catalyze reactions. This is because their models are somehow similar to the models used in computing, and so I think that there is really a possibility of making a major improvement.
Then there are some large-scale problems that can be made more efficient by Quantum Computing. Searching, which is a universal technique, can be sped up by a factor of two, a quadratic factor, by Quantum Computing.
Where I don't think Quantum Computing will be usable is in the typical serial computation that we perform every day.
I have a question about the relevance of these computer systems. Some people might say why not stick to what we have. Why always this push, is it important to have all these new developments and to put forward all this money or would you say there is a point where now it's saturated, we're done.
There are lots of things to discover. This is very important. I think that in Electrical Engineering and in Electronics we are still very far from the optimum solution to problems. So, on the one hand, we should not trust the commercial ventures that tell us to buy their products, as if there was nothing better. Sometimes there is something better than what they can offer in terms of design tools, further down the line.
The other problem is with scientists and academics, because fashions are fashions, and people like to jump from one topic to another, according to how many people listen to them. Very few people actually stick to the fundamental problems, and keep on actually working on those. It's very important that this exercise is done. Like Donald Knuth, a great scientist, who published all these books on the Art of Computer Programming, and made a corpus which is of use to many, many engineers.
You grew up in Italy, you went to Stanford, you went to EPFL, what is on your wish list?
The fact that I spent much of my life in three different part of the world led me to think that there's a home everywhere. I have very good relations with my colleagues in Japan and in China as well as with my American colleagues and South American colleagues, so I don't see any reason why we should build barriers. There's no barrier to knowledge there's no barrier that's useful to the advancement of technology, it is good technology. So I look forward to acting in such a way that, firstly within Europe and then outside Europe, we keep an osmosis of people and technology, because that's beneficial to all.
