Intel/GF deal: Pros, Cons, Unknowns

The industry is still buzzing over a Wall Street Journal report that Intel is in talks to acquire GlobalFoundries (GF) for $30 billion.

It’s been a week since the report appeared. Intel is still mum. GF says there are no talks taking place.

Regardless, it’s worth looking at all of the possible scenarios just in case, and the pros and cons involved. There are layers upon layers of irony behind this would-be deal.

What happens if talks are taking place behind the scenes, and the Intel-GF deal happens? On the flip side, what happens if the negotiations fall through or the companies were never talking in the first place? Here are the scenarios:

Scenario #1: Intel buys GF—Why it makes sense

1) As reported, Intel wants to re-enter into the foundry business after an ill-fated attempt several years ago. By acquiring GF, Intel would quickly gain a large presence in the business. GF is the world’s fourth largest foundry vendor, behind TSMC, Samsung, and UMC, according to TrendForce.

2) There is little or no overlap between Intel and GF. GF mainly focuses on mature nodes. It provides customers with foundry processes at 16nm/14nm, 12nm and above. Then, Intel can focus on the leading edge.

3) Both Intel and GF can collaborate on the further development of advanced packaging. Intel has strong in-house packaging capabilities. GF also has in-house capabilities and has some ties with IBM’s packaging technology. Both see a need to pursue the chiplet strategy.

4) GF is 100% owned by Abu Dhabi’s Mubadala Investment. Since its inception, GF has been swimming in red ink. Mubadala refuses to pour any more money into them. So GF has been talking about an IPO. Intel can sweep in and solve everything.

5) GF is the “Trusted Foundry” for the U.S. In other words, it’s the leading-edge, onshore foundry vendor for the U.S. defense community. GF’s leading-edge capabilities stops at 16nm/14nm finFETs, however. If Intel’s buys GF, the defense community would breathe a sigh of relief. Intel has onshore leading-edge capabilities.

Scenario #2: Intel buys GF—Why it doesn’t make sense

1) GF is a pure-play foundry vendor. It doesn’t compete with its customers. If it buys GF, Intel may end up competing with some of GF’s customers. Those customers may end up shifting orders to other foundries. That defeats the purpose of the deal.

2) Intel has a spotty track record in acquisitions. Big acquisitions are difficult. It’s hard to integrate two large companies. In this case, you have two different corporate cultures—GF and Intel. It’s unclear if the powers that be can make it work.

3) The GF deal fails to address perhaps the biggest problem. Intel remains behind TSMC and Samsung in process technology, and it’s unclear if it will ever catch up. Intel wants to be a foundry player, but it’s unclear if the company will ever become competitive at the high end. In 2018, GF threw in the towel at the 7nm node. No help there.

As reported, Intel was late to the market at the 10nm node, causing it to lose its technology leadership to TSMC and Samsung. Today, Intel is shipping its 10nm chips, but it is now delayed with both 7nm and 5nm. With that in mind, Intel may need to outsource some or all of its leading-edge production to TSMC. Perhaps Intel can deliver its 7nm process. Maybe it will turn out to be competitive. But that doesn’t ensure success in the high-end foundry business. TSMC hates losing orders.

Scenario #3: Intel-GF talks collapse or weren’t real to begin with

1) GF continues to plod along in the foundry business. It will continue to compete in a sizable market. But something needs to give. GF posted a quarter-over-quarter drop of 16% in its revenue to $1.3 billion for the first quarter of 2021, according to TrendForce. GF’s share in the foundry business dropped to 5% in the quarter, down from 7% in the previous quarter, according to the firm.

2) Intel will make another run at the foundry business without GF. Intel may find some business. It will be a niche player at best. The foundry business requires a certain mindset. It’s really a service business. Not sure Intel has the right mindset to keep up against the likes of TSMC, Samsung, UMC, GF, SMIC and others.

3) Even without GF, Intel could put its sizeable resources into the foundry business. The market is ripe for more foundry competition, especially at the leading edge. Pat Gelsinger, Intel’s new CEO, has the right idea. He wants to regain Intel’s past glory. But living in the past is sometimes a risky strategy.