Last week, we reported of a Wall Street Journal post claiming that TSMC was in talks to take control of Intel's fabs in some kind of joint venture. Now it's being reported that is not the case, while yet more reports say that TSMC may indeed take Intel's fabs in a deal that sees Broadcom snap up the chip design side of Intel's business.
Confused? You're not alone. EE Times reports that various analysts have responded to the WSJ report, with one asserting that, "TSMC has no interest in the wafer-fab facilities of Intel."
Meanwhile, another WSJ report doubles down on the idea of TSMC taking over Intel's fabs and adds Broadcom into the mix. The idea is that Broadcom would be interested in acquiring Intel's chip design business if someone else, in this scenario TSMC, takes responsibility for the fabs.
Thus, Intel would both be split in two and cease to exist as an independent entity. If this all seems like wild speculation, it's probably not as crazy as it seems. Currently, Intel's share price is depressed to the extent that its so-called book value, or the value of all its various assets added up, is lower than its market cap, the latter being the total value of all its shares.
In theory, therefore, there's an opportunity to buy a controlling stake in Intel shares, and then break up the company through an asset sale and make a fat profit. In practice, it's rather more complicated than that. However, the fact that Intel's book value is higher than its market cap means it's a nailed-on certainty that all kinds of outfits will be sniffing around Intel's books looking for an opportunity.
No doubt that is why all of these rumours are emerging. Very likely, numerous players are indeed taking a look at a possible deal, either with or without Intel's direct involvement at this stage.
For instance, TSMC is said to be in talks with Intel regarding an injection of engineering nous into Intel's fabs, while Broadcom allegedly has yet to submit any proposals to Intel itself. That makes sense given the complexity of running chips fabs. Intel and TSMC would certainly need to work together on that, meanwhile Broadcom may be able to take ownership of Intel's chip design business with minimal intervention.
Still, there's virtually nothing certain in all of this save for the fact that Intel is clearly in a precarious position after years of struggle and the apparent failure of now-departed CEO Pat Gelsinger's turn-around plan.
For now, all seems to hinge on Intel's upcoming 18A process node. If that's a corker, not only will Intel's own chip designs benefit from a competitive node, but Intel's customer foundry business could finally get some decent clients. In which case, maybe Intel will be reborn.
That would make for one of the great turn arounds in business history. Personally, I think it's more likely that 18A will be somewhere between a bit of a flop and mediocre, I can't see why Gelsinger would have left if 18A really is shaping up to be outstanding.
Thus, it seems odds on that Intel as we known it may cease to exist in the medium term. Whether its split in two, who ends up buying bits of it, really any of the details are impossible to say. But the momentum right now is all in one direction. And it's not Intel as we know it carrying on.
