If things seem to be progressing well on the TSMC side with the new factory in Arizona, despite the disappointment of not finding the latest generation chips there, the rest of the program faces many uncertainties. Indeed, if the two largest parts of it already see their partners Intel and Samsung in a rather significant decline in form, another doubt enters the equation: the passing of the torch to Donald Trump and his team.
Like the report Bloomberga major part of the program Chips and Science Act is already allocated, with 400 billion dollars in state aid already allocated, and 52 billion remaining to be distributed. But everything does not necessarily go smoothly.
There is already the possibility that Donald Trump will decide to revoke current contracts, or even those already concluded. If those which have not yet been completely finalized are the most at risk of being modified or even disappearing completely, those already signed are not necessarily immune: if Joe Biden’s teams affirm that the contracts are made of reinforced concrete , certain clauses could allow a reluctant Trump administration to put spokes in the wheels of this massive plan.
Indeed, the contracts already signed provide in their clauses for performance levels, or for compliance with certain dates. If these levels are not respected by the partner, and this may be due to a simple administrative paper filled out incorrectly or returned 24 hours late, financial sanctions may then apply. This could greatly interest Donald Trump who already thought about the creation of this vast investment plan that it was ineffective, and that import taxes would much more surely motivate companies to come and set up factories in the USA.
Concerns could also be found on the side of the partners chosen to create these factories. If TSMC is in great shape and playing the game, on Samsung’s side the figures are less good, and the company could revise its investments downward. It has already done so by reducing the ambitions of its factory in Texas, and there is nothing to say that the rest will continue as planned.
-But the biggest doubt comes from the biggest contractor: Intel. If the Santa Clara foundry won the biggest bet, with aid of 100 billion dollars, the state of health of the company is more than fragile, and the next manager could very well decide to review the plans in a radically different way compared to those planned by Pat Gelsinger, who was released from the post last year.
If the State has enough to weigh in the balance, with a participation significant enough to influence the decisions taken by the company, nothing says that this weight will be sufficient in relation to financial realities, even taking into account the $3 billion contract that links Intel to the Pentagon.
One solution could be to marry the founder with one of its competitors, GlobalFoundries, especially since the latter is the only one, along with the creator of the x86, to be American. That said, two concerns arise with this alliance: the first, of a technical nature, is that in the end both have difficulties due to the fact that they are unable to maintain the level of their competitors. And as the saying goes, you can’t make a racehorse out of a donkey. The second, of a diplomatic nature: although GlobalFoundries is an American company, it is nevertheless 80% owned by Mubadala Investment Co., the sovereign wealth fund of Abu Dhabi.
The next few months will be decisive for the Chips and Science Act, and more generally the production of more or less advanced chips in the United States. If Donald Trump’s goodwill will weigh heavily in the balance, the financial health of the partners will not be negligible either…
