Breakingviews: Intel’s Tower deal comes up strategically short

Intel has a tower to climb. Buying an Israeli chipmaker of that name is a literal but probably ineffective way to do it.

Once the dominant cutting-edge chip firm, Intel has lost its technology lead. At the same time, the $200 billion company is undergoing a transformation as new-ish Chief Executive Pat Gelsinger attempts to turn it into a contract manufacturer for others’ products. As a statement of intent, Gelsinger’s $5.4 billion acquisition of Tower Semiconductor shows his seriousness about the latter challenge. But with a year or more until closing, Intel can’t bet on this deal to fix its flaws – and Tower could introduce new ones.

True, Tower fits with Gelsinger’s plan to beef up Intel’s manufacturing business. The idea is to make that division sizeable enough to make chips for companies other than Intel itself. The company had also taken a look at $29 billion manufacturer GlobalFoundries, according to the Wall Street Journal. Gelsinger investigated deals that would boost the other side of Intel, too – such as chip designer SiFive, an unlisted rival to industry disruptor Arm. Those talks ended in October, according to Bloomberg.

Intel tried and failed to become a contract manufacturer before, under old boss Brian Krzanich. Bringing in Tower’s management would help to ensure that Intel has a team that knows how to run a foundry. Tower also broadens Intel’s appeal. Many customers don’t need Intel’s latest and greatest manufacturing technology, known as a process node. Instead, they want to use cheaper, older tech of the kind Tower offers. The high investment required to build such capacity from scratch means buying facilities rather than building new ones makes sense.

But Tower’s 22% gross margins badly lag Intel’s. And this is a very small company compared to its acquirer. The transaction also faces a hurdle: China. Chip deals involving U.S. firms can stall for upwards of a year awaiting antitrust approval in the country. In the meantime, existential threats to Intel’s business remain: Taiwan Semiconductor Manufacturing is better at building bleeding-edge chips, and Arm is displacing its designs. Intel needs to spend a lot to catch up. Tower is a nice addition, but it doesn’t much elevate Intel’s prospects.

______________________________________________________________________________________

BREAKINGVIEWS