The smallest objects of utmost importance to the EU

For how small they are individually, microchips play an outsized role in society and geopolitics.

On 21 September 2023, the EU Chips Act came into force, a piece of legislation that aims to "ensure the EU's security of supply, resilience and technological leadership in semiconductor technologies and applications," according to the European Commission.

The act was the result of a call for a common European semiconductor strategy and frees up billions of euros to invest in creating the EU's so-called digital sovereignty — independence from outside producers and vendors of semiconductors.

"In my opinion, we are the most important actor in that strategy," newly appointed Chips Joint Undertaking (JU) executive director Jari Kinaret says.

Chips JU was set up as the first pillar of the EU Chips Act, with the explicit goal to find and invest in semiconductor research and innovation activities, and then connect those efforts to industry. It will manage an expected budget of around €11bn, provided by both EU, Participating States funds and Industry.

Kinaret, a physicist and electric engineer, previously directed the EU-funded Graphene Flagship Project, which successfully managed to create a number of applications and a thriving European industrial ecosystem for the remarkable material.

Barely two weeks into his new role at Chips JU, Kinaret spoke about the main goals, challenges and opportunities for the EU's chips industry.

To start this off, can you walk us through the steps that go into making a microchip?

Jari Kinaret: Now you are challenging my electrical engineering training from a number of years back. But first of all, you start with materials. People used to say that chips are silicon, but if you look at the kind of materials included, just about every material from the periodic table has some role to play in the process.

Then you need to have the design capabilities — you must know what it is that you want to build. That involves, not only the sort of electrical engineering design so that your chip does what you want it to do, but it also includes a lot of physical modelling. It must not get too hot, for instance, and it's not allowed to interfere with other equipment.

Once you have these things in order, then you can actually start building them. You need a facility that typically has a clean room space that is much much cleaner than an operating theatre in any hospital. Then there are a number of very sophisticated pieces of equipment that you need in order to be able to fabricate these structures where the dimensions may be only a few atoms across.\nAnd once you have a functioning chip, you need to protect it from the outside. So you need to handle packaging, which does much more than just protect it from outside nowadays.

Finally, you'd need metrology equipment, because you need to make sure that what you fabricate actually turned out the way that you wanted it to turn out. Because one of the demands is for the yield to be very high — just about 100 percent of the devices that you make must work.

A very complex process indeed. Of this whole production chain, what do and don't we have in the EU?

If we start from what we do have, then we are pretty well positioned when it comes to many of the materials technologies and speciality chemicals that are needed. For instance, to the best of my knowledge, when it comes to the lithography systems, ASML is pretty much the only company in the world that can produce these most advanced systems. When it comes to design tools, there are three main players in the world, one European and two American. And then in the layers in between, like getting the silicon wafers, these are typically from outside Europe. So we are by no means self-sufficient. But neither is anybody else.

So that's where Chips JU comes in

From our point of view, the funding activities earnestly start next year when we start funding a number of pilot lines, aimed at producing new chips. A typical pilot line involves maybe a couple of hundred different pieces of equipment, and some are big, like the extreme ultraviolet lithography systems, some are small, but all are essential.

One of the main motivations for the EU Chips Act is technological sovereignty, as we must be able to continue producing all kinds of smart entities, from cars and aeroplanes to washing machines, even if there are disruptions in the supply chain. So, we need to build our own capacity.

And not all technological capacity, one can imagine. The EU will need skilled workers as well to operate these lines

Part of our funding will go to competence centres, which will provide, if not direct education, at least consulting services that help, in particular, small and medium sized enterprises access the competence that they would be lacking in-house. But I agree that we will, in all likelihood, have to boost science, technology, engineering and maths training so that we can secure the necessary workforce.

And that is mostly done by other parts of the EU funding system, but primarily at national level. So I think that's something that we need to reinforce. And our contribution to that action is mostly through these competence centres, of which there is going to be, hopefully at least one in each EU country.

But in that respect, it's fortunate that the foundries and even the pilot lines do not pop up overnight. It takes a while to build them. So one can start preparing for the shortage of skilled labour.

You've been Director of Chips JU for just a few weeks now, but you must have some plans on what to focus on in the short term

We are busy making sure that we can launch our first funding calls this year, to get moving on the pilot lines in particular. We can't waste any time there.

There are many public and private actors engaged, and we need to act rapidly. In addition, these pilot lines are very complex and diverse, involving lots of different pieces of equipment. It's a rather unique exercise requiring technical expertise and political sensitiveness.

How does this joint undertaking between public and private funding work in practice?

Chips JU comes from the experience gained from predecessors like ECSEL JU, KDT JU and others, bringing together in a tripartite model the European Union, represented by the European Commission, Industry Associations, and Participating States. This apparently complex structure, able to secure a unique leverage effect, allowed us to jointly invest and support hundreds of RD&I projects in the electronic components and systems sector (ECS). Chips JU represents a further evolution of a joint effort to financially support capacity building, as well as innovative projects and ideas for the benefit of the EU and the EU citizens.

The JU will manage different kinds of activities based on a strategic research and innovation agenda, produced by the three European Industry Associations, and a strategic orientation set by the European Commission and Participating States. There is a lot of negotiating involved to reach alignment and effectiveness towards a common goal.

Speaking of funding, €11bn is a huge amount of money. But considering the fact that a new foundry costs upwards of €20bn, it suddenly seem piddling

On a human scale it's a very large amount of money. But compared to the current and future needs at European and global level it's not.

We are aware of huge investments made in other geographies. The magnitude of investment in order to be a part of the global competition in the semiconductor sector is not easily achievable, that's why a joint effort is needed. Nonetheless, the Chips Act mobilises an unprecedented amount (43bn euros) and we have been asked, as Chips JU, to implement an important part of it, making us a significant actor in the semiconductor ecosystem.

Just out of curiosity, what are some developments when it comes to the production of chips that you're personally excited about?

There are a lot of new ideas and, amongst them, I would like to mention, for example, neuromorphic computing. This method would allow computations to resemble more the way the brain computes than what today's computers do.

Also, because of my past, I see that many two-dimensional materials, not necessarily graphene itself, but many others, like molybdenum, disulfide, and others show great promise. I'm convinced that they will find their way to mainstream electronics, maybe not in the next five years, but more likely in the next 10 years.

The Chips Act also sets forward some goals for the future. Where do you see the EU stand in terms of technological sovereignty in 2030?

I am convinced that our position will be much stronger than what it is today. Today about 10 percent of chip production takes place in Europe and the target for 2030 is 20 percent. This is a very ambitious target because the market size is expected to double during that period, which means that our capacity should quadruple.

This also involves all kinds of strategic choices. As we said about our budget, it's large on a human scale, but still small on the scale of needs. We need to be conscious that we cannot support priorities at the same pace and with the same amount of funding. And we need to decide where we should put our money first. This is a political choice that needs to be taken together with industrial actors. The direction will also be influenced by commercial choices.

To mark the Chips Act adoption and the Chips JU establishment, on 30 November and 1 December, we will offer to political and industrial representatives the opportunity to meet and exchange their views at the Chips for Europe event, a public gathering taking place in Brussels. With the presence of Commissioner Breton, MEPs, CEOs, CTOs, SMEs, RTOs, and high-level public authority figures will deliver crucial political and industrial messages regarding the Chips Act and Chips JU's ambitions. I would like to take this opportunity to invite you and your readers to join the web streaming of the event and be part of this transformative chapter in the world of microelectronics!

Follow the live stream of the Chips for Europe event on 30 November and 1 December.