What is SiPearl's DNA? What are you selling? What are you planning to sell tomorrow?
SiPearl designs very high-performance electronic components and microprocessors used in supercomputers and the development of artificial intelligence. We design, but we don't manufacture. Like many companies in the sector, we are 'fabless.' This strategy is driven by the massive costs (billions of dollars) of the factories that need to be pooled, as well as the technical expertise and time required to set up a factory. We are currently developing our first generation of semiconductors, which will be produced in Taiwan, but this production in Taiwan in no way weakens our European DNA.
SiPearl emerged from our vision to establish a European benchmark in the semiconductor sector, focused on the high-performance computing segment. Our mission resonated with European authorities, who recognized its significance and supported our efforts with essential funding. They have also opened doors for us to work on European projects such as the German supercomputer Jupiter. Our teams are spread across seven centres in Europe and share our in-house culture, which aims to provide the technical tools for greater European sovereignty. Even British engineers, with whom we've engaged in discussions, understand and share our commitment to technological sovereignty, especially in relation to the United States. SiPearl's foundations are robust, and our ambitions are both significant and well-founded.
How does a young startup like SiPearl plan to establish a significant presence in the highly competitive microprocessor market?
We are a young start-up, it's true, but we have technological advantages that already set us apart from many of our competitors. Our strengths include architecture and energy savings. When it comes to security and protecting our users, we have a clear competitive advantage: there are no hardware or software backdoors in our products, and all our components can be audited to reassure users. This issue is increasingly relevant today, and we must respond to these legitimate expectations.
Our aim is to give Europe and all our users greater sovereignty, particularly in relation to the United States while building a leading position in the global application markets of artificial intelligence and supercomputers. This European sovereignty can be embraced by other countries outside Europe because we share the same values of independence.
What are SiPearl's objectives for the next 5 or 10 years?
Our goal over the next 5–10 years is to become the European leader in supercomputing. We already have everything it takes to become the most advanced company in all the components of supercomputing. Although it will be difficult to reach a higher level than the Americans in this timeframe, I think we can reach the same level if the willingness—particularly on the part of European politicians—is there. Ultimately, we have the potential to become a company like STM was 25 years ago, ranked among the top five globally, albeit in a different sector than ours.
Can Europe still maintain its sovereignty in the face of global giants such as the USA and China?
There is no ready-made answer to this question. It will depend on Europe's ability to mobilise and make the right decisions in the long term. First of all, I believe it is essential to put in place a 'European Buy Act.' Our funding, our R&D, and our industries must be much more at the service of Europeans. In addition, we need the political willingness to fight everything to do with American extraterritoriality. The Chinese are beginning to follow the same logic. Europeans, however, tend to be far more naïve and should be taking similar actions. This is quite frustrating, particularly as only political measures can effectively address this issue.
However, the European Chips Act is a good start. If we manage to mobilise more public funds in the semiconductor sector to get things moving again, as is being done in most countries, that will be a positive thing. Nevertheless, we need to take a close look at who benefits from this funding. It is understandable that large groups such as Infineon, Bosch, STM, and ASML should benefit from this windfall, but the problem is that start-ups have been left by the wayside. In other words, the Chips Act is poorly targeted and nearly ended up benefiting not Europeans but Americans. Intel, for example, proposed a $10 billion investment in its German site in Magdeburg to produce some of its components. An interesting strategy for Intel, which would have ensured that around 70–80% of its production capacity would be mobilised. But why should European public funds enrich an American company? In the end, and fortunately, the operation did not go ahead. The Chips Act addresses some market needs, but unfortunately, it falls far short. The system is still not well-designed enough to support the emergence of the next Nvidia on European soil.
Are geopolitical tensions, especially those involving Taiwan, a potential vulnerability for SiPearl's development?
I would rather talk about a risk than a weakness with our microprocessors produced in Taiwan. Taiwan is a must, producing the best semiconductors in the world for all the companies in the sector. The authorities view Taiwan's unparalleled capabilities as the best safeguard against China's ambitions. If, by any chance, the situation in Taiwan were to significantly deteriorate, all the players in the semiconductor industry would be severely affected.
The Taiwanese company TSMC has a plant with significant capacity in the United States that could become an alternative solution, and another is due to open in Germany, but using less advanced technology than in Taiwan. Samsung has production units in South Korea and the United States. Not everything is concentrated in Taiwan, but we certainly need to remain vigilant in the face of major geopolitical movements. In an ideal world, we would manufacture in Europe, but the current situation compels us to make strategic choices. While there is a geopolitical risk with Taiwan, we can at least be confident that everything will operate smoothly from a technical standpoint.
Numerous microprocessor factory projects are underway in Europe, but major delays or cancellations have been announced in the last months. Isn't this worrying?
This situation seems perfectly logical to me, first and foremost, because there is no market. Building a plant with advanced capabilities in Europe is all the more difficult because there is no product, no start-up, and no financing behind it. There is a high risk that the plant will be under-operated.
Having the determination to set up factories is a necessary step, but it does not guarantee the sustainability of the approach. This is all the more true because engineers and technicians are needed to build and run a plant. There is a real shortage in this sector, affecting both Europe and the United States. The TSMC plant in Arizona (USA) is still experiencing difficulties because the skill levels and work capacity of the Taiwanese are not matched by their American counterparts. By 2030, there is a risk that resources will be saturated. Our industry also has major needs for water, electricity and, of course, technicians and engineers, three 'ingredients' that money alone cannot buy in the short term. The challenges ahead are, therefore, immense and multi-faceted. Anticipation and adaptability will be essential for companies aiming to play a major role in the future of microprocessors and, by extension, the global economy.
