Intel's takeover dilemma: A Gordian knot of funding and politics

Jay Goldberg

Posts: 95   +2
Staff
Editor's take: A takeover of Intel has become a Gordian knot. The company's factories require massive investments – billions of dollars and several years to fix – which most prospective buyers, companies or private equity firms aren't willing to handle. The US government has also poured a lot of money into these factories, making it politically difficult to shut them down. The problem is clear: no one wants the factories, but Intel can't be sold without them.

No one wants to buy Intel's fabs and their bottomless funding needs, but the company probably cannot be sold without someone taking them on.

There's been more stories this past couple of weeks about Intel getting acquired. An industry analyst claimed that he "had been read a letter" which showed some unnamed company was in the process of buying the company. We are not going to link to that report because we are not certain about its accuracy... something in the language feels "off" to us. Nonetheless, it sparked a surge in the company's share price and a host of follow-up coverage.

Editor's Note:
Guest author Jonathan Goldberg is the founder of D2D Advisory, a multi-functional consulting firm. Jonathan has developed growth strategies and alliances for companies in the mobile, networking, gaming, and software industries.

This is not exactly news – Reuters reported this back in September, and to be blunt, we trust Reuters' sourcing a lot more. Going back even further, we speculated that Broadcom could buy Intel all the way back in May before things got really bad at Intel. We were only semi-serious about that idea, but this timeline has clearly entered reality-is-stranger-than-fiction territory, and it looks likely that Broadcom at least considered a bid. As we have been saying, Intel is in play, and anything is possible.

Well, not anything. There are only a handful of companies capable of buying Intel, and the two most likely candidates have already taken a look and seem to have stepped away. There are also a dozen or so private equity funds who could afford it, but our sense is that they are passing on it as well.

A takeover of Intel has become a Gordian knot. The big problem is funding the company's fabs, which will require tens of billions of dollars and years to get back on track. Few companies, and no private equity funds, really want to deal with that large of a funding need and time horizon. On the other hand, the US government has given Intel a lot of money, and so simply shutting down the fabs is deeply problematic. No one wants the fabs, but the company cannot be sold without them.

In theory, the new administration could give a buyer approval to shut the fabs, but if someone has enough political capital for that purpose, why not use that political capital to secure some direct government support?

In speaking with investors, our impression is that the Street assumes the only way to save Intel is for the government to intervene. We maintain that this is not a hard requirement, but we recognize that this is now the common perception of the situation. There are, of course, rumors that a certain highly-connected, deeply troubling tech mogul has a plan to buy the company. And from the very narrow perspective of saving the US' semiconductor manufacturing capacity, that may be what it takes.

We are increasingly convinced that the only way for Intel to survive is for someone to buy them and remove the board. Unfortunately for the company, and the semiconductor industry, that path looks very challenging.

Permalink to story:

 
Intel resting on its laurels was its undoing. The prolonged reliance on 14nm was a clear sign of stagnation, and moving beyond it proved to be an uphill battle. Meanwhile, AMD, though not dominant for much of the last decade, managed to innovate just enough to gain significant traction and continues. This, in turn, exacerbated Intel's challenges.

The competition with TSMC highlights Intel's deeper struggles. While TSMC continues to advance its process technology at a steady pace, Intel's own fabrication business remains its Achilles’ heel.

For Intel to regain its edge, it must refocus on innovation. However, with the current leadership and organizational hurdles, this may be easier said than done. Without bold changes, Intel risks falling further behind in an industry that thrives on relentless progress.
 
Best solution for us all: Samsung gets the fab branch heavily discounted.
They've got the experience and expertise as they're TSMCs closest and only competitor for high end processes if Intel goes the way of the dodo. Getting all of Intels intellectual property and combining it with their own might allow them to compete with TSMC a lot better.

Get the US government to throw in some billions (which would be a bit of a struggle as Samsung is a South Korean - not American company) and perhaps assist with brokering a deal with NVIDIA and/or Samsung to be garantueed customers so they know the fabs will keep running for at least a while. Perfect candidate for the PS6/XBox whatever chips and some of the lower positioned desktop GPUs/CPUs.

Alternatively, NVIDIA's market cap is stupid high at the moment and drowning in orders they can't get out of the door for years. Getting a whole lot of production capability would be benificial if TSMC was somehow open to licensing their processes or NVIDIAs customers can settle for slightly worse energy effiency.
NVIDIA being the company it is this would probably be terrible for the market if successful.

Whichever way it goes, the US would be stupid to sink billions into domestic chip capabilities just to let their biggest actually domestic manufacturer fabs come to a standstill. Trump would probably have had some more wiggle room financially to make something happen if project Starfall and it's half a trillion budget wasn't just announced.
 
Amusingly, TSMC competing with Intel is only via AMD. Otherwise the PC market could've pondered on as it was.
And Intel can't "innovate" its fabs further without its gluttonous revenue of old. Well, not until it trims a lot of fat that is. It needs to find profits again.
 
Amusingly, TSMC competing with Intel is only via AMD. Otherwise the PC market could've pondered on as it was.
And Intel can't "innovate" its fabs further without its gluttonous revenue of old. Well, not until it trims a lot of fat that is. It needs to find profits again.
TSMC competes with Intel via Apple, considering Apple used to be an Intel customer.
 
Intel resting on its laurels was its undoing. The prolonged reliance on 14nm was a clear sign of stagnation, and moving beyond it proved to be an uphill battle. Meanwhile, AMD, though not dominant for much of the last decade, managed to innovate just enough to gain significant traction and continues. This, in turn, exacerbated Intel's challenges.

Intel was stuck at 14nm basically forever because they thought they could do 10nm without having to invest in EUV. And yes, the fact AMD was not a threat during that timespan (I think that was Bulldozer/Piledriver) probably aided in that perception. Throw in Intel basically continuing to rehash a CPU design that dates back to Sandy Bridge, and is it any shock things started to fall apart in short order?

What I would do is *not* nationalize Intel, but I would nationalize their fabs, and open them up to whoever wants to use them (barring some national need in which case, sucks for consumers). That's the only way to untangle the knot.
 
Broadcom buying Intel would be wild, but honestly, it feels like the real player here might end up being the US government. Between CHIPS Act funding and national security concerns, Intel’s future is increasingly tied to politics rather than just the market.
 
Amusingly, TSMC competing with Intel is only via AMD. Otherwise the PC market could've pondered on as it was.
And Intel can't "innovate" its fabs further without its gluttonous revenue of old. Well, not until it trims a lot of fat that is. It needs to find profits again.
Your a little off, TSMC customers are Apple first and foremost, their biggest client for all their devices.

AMD
Nvidia
Qualcomm
Broadcom
Sony
Marvell
Tesla
Xilinx
STMicroelectronics
Texas Instruments and a few other strays I cant seem to recollect.

Current market cap is ~$989B
 
Broadcom buying Intel would be wild, but honestly, it feels like the real player here might end up being the US government. Between CHIPS Act funding and national security concerns, Intel’s future is increasingly tied to politics rather than just the market.
You’re absolutely right!
It would be a seismic move, but the geopolitics tied to semiconductor manufacturing makes any potential deal way more complex than a typical corporate acquisition. With the CHIPS Act pushing for domestic manufacturing and reducing reliance on foreign chip production, Intel is positioned as a cornerstone of the U.S.'s tech sovereignty strategy.

Broadcom, while American, operates heavily in networking and data center markets but doesn’t have the same symbolic status as Intel in the U.S.'s national tech agenda. A merger would raise major flags, not just over market consolidation but also regarding supply chain security, R&D, and defense applications.

Do you think regulators would even let something like that pass, given how closely Intel is tied to U.S. government interests?
 
I sometimes wonder if Berkshire Hathaway would be interested in Intel. BRK have something like $250 billion in cash sitting there, they are generally pro-US industry and love a cheap blue chip stock. They do also love a cheap, profitable stock which intel definitely isn't currently, but that can be priced into the buy-out price as well. A BRK buyout would also present minimal resistance from regulators, unlike a Broadcomm, Apple or Nvidia buyout.

I haven't heard any rumours of that yet, but it is an interesting thought experiment at least.
 
I sometimes wonder if Berkshire Hathaway would be interested in Intel. BRK have something like $250 billion in cash sitting there, they are generally pro-US industry and love a cheap blue chip stock. They do also love a cheap, profitable stock which intel definitely isn't currently, but that can be priced into the buy-out price as well. A BRK buyout would also present minimal resistance from regulators, unlike a Broadcomm, Apple or Nvidia buyout.

I haven't heard any rumours of that yet, but it is an interesting thought experiment at least.
This thought actually entered into my head a while back when this all started.
Berkshire/Hathaway have their hands in just about everything.
 
Broadcom, while American, operates heavily in networking and data center markets but doesn’t have the same symbolic status as Intel in the U.S.'s national tech agenda. A merger would raise major flags, not just over market consolidation but also regarding supply chain security, R&D, and defense applications.

Do you think regulators would even let something like that pass, given how closely Intel is tied to U.S. government interests?

Based on how Broadcom handled the VMware acquisition, my guess is that would be looked at with scrutiny. But then, there aren't just economic factors at play here, and the political regime has just changed hands.

I'm not entirely convinced that Intel needs to be bought out. While this article made it plain that Intel needs money and investments in its fabs, it didn't make the case that Intel can't get that funding from the government, or that temporarily relying on TSMC for its manufacturing (like it is doing for Arc and Battlemage) while Intel gets its fabs in order would be a bad option, either. Intel may not have what TSMC has right now, but there are plenty of technologies that depend on older (and cheaper) process nodes - Intel could surely make chips for those markets, too. Easier said than done, I'm sure.

There are plenty of companies rich enough to buy Intel, but so far choosing not to. In fact, I feel like Intel's problems, while significant, are nothing compared to what Boeing is going through right now (at least Intel doesn't have blood on its hands, to my knowledge), so I for one am not as pessimistic as this article is.
 
Intel resting on its laurels was its undoing. The prolonged reliance on 14nm was a clear sign of stagnation, and moving beyond it proved to be an uphill battle. Meanwhile, AMD, though not dominant for much of the last decade, managed to innovate just enough to gain significant traction and continues. This, in turn, exacerbated Intel's challenges.

The competition with TSMC highlights Intel's deeper struggles. While TSMC continues to advance its process technology at a steady pace, Intel's own fabrication business remains its Achilles’ heel.

For Intel to regain its edge, it must refocus on innovation. However, with the current leadership and organizational hurdles, this may be easier said than done. Without bold changes, Intel risks falling further behind in an industry that thrives on relentless progress.
Intel wasn't on 14nm because they wanted to be there, they didn't rely on 14nm, their fab couldn't keep up with their designs, due to how difficult a node shrink is. Manufacturing CPUs isn't the same as designing them.
 
Intel wasn't on 14nm because they wanted to be there, they didn't rely on 14nm, their fab couldn't keep up with their designs, due to how difficult a node shrink is. Manufacturing CPUs isn't the same as designing them.
Intel is Intel. Whether it's fabs or design, the result is the same, they let AMD advance.

If node shrinks were that difficult, TSMC wouldn’t be pulling them off consistently.
 
Back