Uphill race for chips in Europe

Three new factories under construction in Europe. Fifteen in the rest of the world. To the question of whether it is possible to significantly increase the market share of chips produced in Europe an initial answer is here, and it is clearly a negative one. The target set by the European Commission of 20% of world production of microprocessors by 2030 seems totally unrealistic, and it is therefore preferable to look at other avenues, so as to improve the efficiency of processes and speed up the time-to-market of individual products, not least through the use of the new tools made available by artificial intelligence. This is the ultimate sense of a paper by Porsche Consulting on the subject, produced by first of all lining up the existing numbers, which are not very reassuring for the European Union. Against Chinese public funds of almost 150 billion, the EU budget is limited to 46 billion, surpassed even by South Korea. The market share target of 20% in the production of microprocessors seems distant, "disconnected from reality", the report explains, citing the opinion of independent audits on the sector: if in 2022 the European share was just under 10%, the projections indicate for 2030 the possibility of arriving a little further, at around 12%-13%, thus leaving a significant gap to be bridged. How? Porsche Consulting's recipes look first and foremost at a recovery of efficiency, with the need also to change some rules of the game if progress is to be made more quickly. 'Given the vastness of the areas of use, with an increasingly pervasive spread of electronics,' explains Porsche Consulting partner Giovanni Notarnicola, 'total self-sufficiency in this area will be impossible for any area of the world, no one will be able to be completely decoupled from the others. In Europe, however, there are additional obstacles that need to be removed, starting with the cost of energy, which is up to 2-3 times higher than in other areas of the world. This severely limits the competitiveness of continental production and thus also the ability to attract new investments to the area'.that remain crucial to prepare for the future, taking into account the long lead times for the commissioning of each plant. The Esmc consortium in Germany, for example, (Tsmc, Bosch, Infineon, Nxp) has seen the start-up of the joint venture in 2023, with the first wafer productions planned in reduced quantities (40,000 per month) only in 2027, to then scale up to almost 500,000 units per month in 2029, thus six years after the company's first start-up.

'Given the times,' Notarnicola explains, 'we also need to look elsewhere, at boosting the productivity of existing sites and improving the time to market of products. Artificial Intelligence can play an important role and this is where companies must invest, also aiming to 'virtualise' research and development activities as much as possible, thus reducing the time and costs of the activity'.

The estimate on a single product,' the paper explains, 'is to reduce costs by $50 million and cut development time by one year.