IBM's help...
So IBM is going to show them how NOT to do it?
Japan looks set inject 70 billion yen ($500 million) into a new semiconductor company amid plans to jump-start next generation advanced chip production with the help of IBM. Earlier this week, we reported on the nation’s preparations for a joint research program with the US to build cutting edge chip plants. The project will …
IBM isn't in there for its (lack of) business acumen, but for its RnD.
As the article states, IBM was the first to make 7nm, 5nm and 2nm wafers.
In semiconductor RnD, IBM has many firsts, and its technology is top-notch, one of their staff even got a Noble Prize for their work - while at IBM - on semiconductors. It's their ability to do anything useful with that tech that's questionable.
It's going to involve 1000s of meetings where everyone is very polite and calls everyone '-san' but nothing is done because it has already been decided that it isn't going to happen, but everyone is waiting for the other side to drop out first to save face.
More than a dozen Japanese organisations are involved, so there are factorial(12) combinations of this.
All the organisations (including IBM) are basically in managed decline, so they are all just waiting this out for grants and to be seen to be doing something by their superiors
How do you set up a semiconductor business with $500 m? That sounds like a mighty small amount of money.
Gov officials in this place here where I stay will steal an equivalent amount of money in about 2 weeks for no benefit to anyone beside themselves.
It is interesting what IBM is doing. They sold off their foundry business to Global Foundries when they decided it was uneconomic. Then they got bent out of shape when Global Foundries decided that developing nodes smaller than 14nm was uneconomic. Now they are having a go with the Japanese. Which is interesting, because the Japanese had decided that keeping up with node reduction was uneconomic. Something tells me they will be buying from TSMC in the near future.