Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/tsmc-considers-running-intel%E2%80%99s-us-factories-after-trump-team-request.22101/
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021770
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

TSMC Considers Running Intel’s US Factories After Trump Team Request

XYang2023

Well-known member
(Bloomberg) -- Taiwan Semiconductor Manufacturing Co. is considering taking a controlling stake in Intel Corp.’s factories at the request of Trump administration officials, a person familiar with the matter said, as the president looks to boost American manufacturing and maintain US leadership in critical technologies.

Trump officials raised the idea of a deal between the two companies in recent meetings with the Taiwanese chipmaker, the person said, and TSMC was receptive. It’s unclear whether Intel is open to a transaction.

The talks are in very early stages, and the exact structure of a potential partnership hasn’t been established. But the intended result would have the world’s largest made-to-order chipmaker fully operating Intel’s US semiconductor factories, said the person, who asked not to be identified because the conversations are private. It also would address concerns about Intel’s deteriorating financial state, which has forced the company to slash jobs and curb its global expansion plans.

The arrangement may involve having major American chip designers take equity stakes, according to the person, along with support from the US government. That means the venture wouldn’t solely be owned by a foreign company. TSMC is the go-to chipmaker for Apple Inc., Nvidia Corp. and other companies developing semiconductors that power AI algorithms.

Still, the possible partnership could run into political hurdles, not unlike those that have hamstrung a proposed acquisition of United States Steel Corp. by Japanese maker Nippon Steel Corp.

A White House official said that the president is unlikely to support a foreign entity operating Intel’s factories.

TSMC and Santa Clara, California-based Intel declined to comment.

Intel shares pared losses on Friday after Bloomberg reported on the discussions. The stock was down 2.2% to $23.60 at the close in New York after dropping as much as 5.3% earlier in the session.

While Intel has lost a lot of ground to rivals over the last five years, it still manufactures the most widely used components in the personal computer and server industries. And it has the largest and most advanced production network owned by a US-based company. Those factories have taken on vital strategic importance as Washington works to reverse decades of semiconductor manufacturing shifting to Asia.

Under former Chief Executive Officer Pat Gelsinger, Intel embarked on an ambitious and expensive plan to restore its chipmaking lead, and the company won $7.9 billion in US government funding to support projects in four states. It’s also secured $3 billion to produce chips for the US military, all of which will be paid out over time as Intel’s plants hit key milestones. The company has received $2.2 billion as of January.
But that effort has so far failed to attract enough outside customers to make the investments worthwhile, particularly at a new site in Ohio. Intel’s own products also are losing market share, adding to the overall squeeze on its finances — just when it needs to spend heavily. Gelsinger was forced out in December after the board lost confidence in his turnaround plans.

Though Intel has reiterated its commitment to investing in US factories — even as it pulls back from efforts abroad — Washington officials have for months been discussing contingency plans. One option favored by some in the Biden administration was a possible deal between Intel and GlobalFoundries Inc., Bloomberg has reported.

That would have married a maker of older-generation chips with Intel’s leading-edge business. But GlobalFoundries walked away from cutting-edge production years ago and doesn’t have the money for an acquisition, so those conversations didn’t progress much past a thought exercise.
Biden officials also floated the possibility of TSMC licensing its manufacturing technology for use at Intel’s facilities, according to multiple people familiar with the conversations. But TSMC wasn’t interested in an arrangement that could ultimately benefit a competitor, the people said. And Biden’s team was generally reluctant to take an active role in deal conversations, according to several people familiar with the situation.

Trump, meanwhile, is not shy about cutting deals — and TSMC, which just held its board meeting in the US for the first time, is eager to maintain good relations with the new president.

Trump has repeatedly accused Taiwan of “stealing” the US chip industry and threatened tariffs on foreign-produced semiconductors. He’s expressed a preference for using those kinds of levies to boost US chipmaking instead of government subsidies — the approach adopted by the US Chips and Science Act. That legislation, passed in 2022, led to TSMC winning $6.6 billion in grants to support three plants in Phoenix.

But it was also Trump’s first administration that set out to lure TSMC to the US, and top Trump officials have consistently affirmed their commitment to boosting domestic manufacturing. That’s particularly true for technologies at the heart of the US-China competition.

The US “will ensure that the most powerful AI systems are built in the US with American-designed and manufactured chips,” Vice President JD Vance said during an AI summit in Paris earlier this week.

The fact that TSMC is open to an Intel deal indicates a change in tune from just a few months ago. In October, CEO C.C. Wei had told analysts that his company wasn’t interested in acquiring Intel’s factories.

Asked the same question again in January, Wei didn’t directly answer: “They are our very good customers,” he said. “I like them and they are very important to TSMC’s business also. That all I can say.”

 
If this story is true, Trump administration is trying to arrange a deal to kill Intel, the Intel we know today.

Let's call it "forced assistant suicide".
It almost seems like a shame. The Intel board should not have fired Gelsinger before delivering on 18A to allow people to decide whether Intel should be killed off or not.
 
Even if this is true, it isn't like Intel is going to stop their 18A node given that their newer processors they're producing later in the year are using it (i.e., Panther Lake...).
 
This won’t be the first time TSMC was approached. The last time it happened the CEO was fired.

Given the manufacturing history and culture of Intel as well as how the Fabs in Taiwan and Fab21 and WaferTech are run barring a complete house cleaning of management operationally a merger is laughable. How it is run down to how they even review problems could is so different in expectation. Can a zebra change their strips? I’d say 90% of the Intel mangers to the highest level couldn’t survive the daily expectations of TSMC.

As to if this involved some strategic technical help also laughable. TSMC barely can find the talent to fill and manage the US as well as the expansions in Taiwan, Japan. Now they got Germany too. Taking valuable talent into a totally different organization culture and way to do business to try to solve the technical as well as cultural and organizational will be laughable too in probability of success and a waste of valuable TSMC futility resources.

Leading edge technology is the most complex technical and operational thing on earth. To think bringing say ten to twenty or even a hundred key people can fix something is ludicrous.

Intels problem isn’t the technology or factory, it’s the far larger business and strategic approach to foundry and how they dance with customers, develop and deploy PDK and how that is managed and executed. To think TSMC can swoop in and managed that to success, right
 
This won’t be the first time TSMC was approached. The last time it happened the CEO was fired.

Given the manufacturing history and culture of Intel as well as how the Fabs in Taiwan and Fab21 and WaferTech are run barring a complete house cleaning of management operationally a merger is laughable. How it is run down to how they even review problems could is so different in expectation. Can a zebra change their strips? I’d say 90% of the Intel mangers to the highest level couldn’t survive the daily expectations of TSMC.

As to if this involved some strategic technical help also laughable. TSMC barely can find the talent to fill and manage the US as well as the expansions in Taiwan, Japan. Now they got Germany too. Taking valuable talent into a totally different organization culture and way to do business to try to solve the technical as well as cultural and organizational will be laughable too in probability of success and a waste of valuable TSMC futility resources.

Leading edge technology is the most complex technical and operational thing on earth. To think bringing say ten to twenty or even a hundred key people can fix something is ludicrous.

Intels problem isn’t the technology or factory, it’s the far larger business and strategic approach to foundry and how they dance with customers, develop and deploy PDK and how that is managed and executed. To think TSMC can swoop in and managed that to success, right
Well said.
 
TSM rules the high end and will continue to do so. Cutting a deal with the largest power on earth is just common sense. It won't be easy, but if handled properly, both sides could win. A win/win situation is almost always preferable to conflict. If done properly, both sides could win big time with both sides and the world benefitting. Solving the corporate culture issues will be the biggest challenge, but the rewards could be literally staggering for both sides if they could meet the challenge.
 
TSMC should look hard at any deals to expand manufacturing rapidly in the US.

TSMC according to Morningstar has a price to fair value of 0.75. This is a major, major problem, as bad as what got Gelsinger fired.

The market discounts TSMC stock heavily over geopolitical risks.

I think P/FV will rise if TSMC management can find a way to expand rapidly outside of TW. They are working on this. There are now TSMC fabs in Japan and Arizona, and soon in Germany. But this isn't rapid, it's cautious (as it should be I think).

Imagine TSMC stock price with a Nvidia multiple. It could be $1000. A deal to expand rapidly outside of TW can make that happen.

The Irish cement firm CRH decided to expand in the US and it boosted their multiple. In fact it's a bit of a case study in how investing in the "renewed US growth" story was a smart move.

For decades, until 2022, Foreign direct investment was all directed toward China, on the assumption of superior future growth. But China's original growth thesis is spent. And the US has a new growth thesis: AI. So investing in the US to get some of the growth, and a better multiple, is something all global CEOs should be considering, TSMC included.
 
It almost seems like a shame. The Intel board should not have fired Gelsinger before delivering on 18A to allow people to decide whether Intel should be killed off or not.
The rumor was that he was fired for having an AI chip roadmap that was uncompetitive. Considering that Falcon Shores is now no longer a commercial product, seems like there might even be teeth to that rumor. A real rarity among most of the Intel ""news"".
TSM rules the high end and will continue to do so. Cutting a deal with the largest power on earth is just common sense. It won't be easy, but if handled properly, both sides could win. A win/win situation is almost always preferable to conflict. If done properly, both sides could win big time with both sides and the world benefitting. Solving the corporate culture issues will be the biggest challenge, but the rewards could be literally staggering for both sides if they could meet the challenge.
I disagree. Besides having no reason to not expect TSMC to bungle it, it seems like a bad deal. TSMC has to put lots of money effort and time to maybe turn around the business and completely screw up their capacity LRP. Intel's legacy fabs can't run new processes and TSMC utilization for their legacy processes is subpar with even more fabs coming up in Japan and Germany. So an Intel merger doesn't really save TSMC capex dollars. Intel's EUV fabs are mostly scattered around and while they add up to a good chunk of capacity TSMC'S customers won't accept chips from say Ireland if they did their qualification in Arizona. The biggest benefit to TSMC is killing their most serious competitor since the very early 2000s in the crib before they've had time to prove their capability (or maybe lack thereof).

Since there are already internal and external foundry products designed for 3, 18A and 14A TSMC can't really convert any Intel fabs to TSMC technology for years until the existing production runs start winding dow. So if they really wanted to avoid tariffs they should just accelerate their AZ build out and look at another Greenfield site near Dallas, Boise, Austin, or upstate NY (assuming the local parties are willing to cut a competitive deal my pick would be Dallas given the large TI presence to draw local experience from). That approach should be faster and at the very least cheaper since you aren't paying for any of the assets you don't want.

Also seems like a rotten deal for Intel too since progress seems solid against Intel foundry's financial and technical deliverables (at least based on Intel guidance or any reasonable person's understanding that signs of recovery cannot start to show up until 2026 at the earliest due to long chip design cycles). Why give up after investing tons of cash over the past 3.5 years just before you start generating returns? For two years now Intel said 2024 and especially 2025 would be the years where capex and losses would peak. Any deal TSMC would cut would be a lot smaller than the potential upside, lower Intel's gross margins, and lower their maximum operating margin potential beyond 2026. The market is ripe for sizing with Samsung's repeated failures, TSMC slowing down, and tensions of their leading edge monopoly seemingly rising with TSMC's ever growing cut of the pie (margins).

All in all seems like a lose-lose rather than a win-win.
 
The rumor was that he was fired for having an AI chip roadmap that was uncompetitive. Considering that Falcon Shores is now no longer a commercial product, seems like there might even be teeth to that rumor. A real rarity among most of the Intel ""news"".

I disagree. Besides having no reason to not expect TSMC to bungle it, it seems like a bad deal. TSMC has to put lots of money effort and time to maybe turn around the business and completely screw up their capacity LRP. Intel's legacy fabs can't run new processes and TSMC utilization for their legacy processes is subpar with even more fabs coming up in Japan and Germany. So an Intel merger doesn't really save TSMC capex dollars. Intel's EUV fabs are mostly scattered around and while they add up to a good chunk of capacity TSMC'S customers won't accept chips from say Ireland if they did their qualification in Arizona. The biggest benefit to TSMC is killing their most serious competitor since the very early 2000s in the crib before they've had time to prove their capability (or maybe lack thereof).

Since there are already internal and external foundry products designed for 3, 18A and 14A TSMC can't really convert any Intel fabs to TSMC technology for years until the existing production runs start winding dow. So if they really wanted to avoid tariffs they should just accelerate their AZ build out and look at another Greenfield site near Dallas, Boise, Austin, or upstate NY (assuming the local parties are willing to cut a competitive deal my pick would be Dallas given the large TI presence to draw local experience from). That approach should be faster and at the very least cheaper since you aren't paying for any of the assets you don't want.

Also seems like a rotten deal for Intel too since progress seems solid against Intel foundry's financial and technical deliverables (at least based on Intel guidance or any reasonable person's understanding that signs of recovery cannot start to show up until 2026 at the earliest due to long chip design cycles). Why give up after investing tons of cash over the past 3.5 years just before you start generating returns? For two years now Intel said 2024 and especially 2025 would be the years where capex and losses would peak. Any deal TSMC would cut would be a lot smaller than the potential upside, lower Intel's gross margins, and lower their maximum operating margin potential beyond 2026. The market is ripe for sizing with Samsung's repeated failures, TSMC slowing down, and tensions of their leading edge monopoly seemingly rising with TSMC's ever growing cut of the pie (margins).

All in all seems like a lose-lose rather than a win-win.
I wish you could tell Frank Yeary, Intel’s chairman, all that. He seems to be pushing for this based on this well sourced NYTimes article. One upside of pushing this is that the all the laid off Intel development engineers will help drive down wages for semiconductor engineers in the US and make things more cost competitive.

 
The rumor was that he was fired for having an AI chip roadmap that was uncompetitive. Considering that Falcon Shores is now no longer a commercial product, seems like there might even be teeth to that rumor. A real rarity among most of the Intel ""news"".
Hard to be fired for something you inherited. No I think one of several

1). A whale client said to do a deal he had to go

2). Board wanted to push on a strategy change and it was either get on board or leave.

I am thinking the second as things are pretty dire and require a pivot. They wasted tens of billions on shells that they didn’t need. Foundry should build to real demand not hope and prayer. That is a spectacular failure of Pat’s part also somewhat the board and CFO
I disagree. Besides having no reason to not expect TSMC to bungle it, it seems like a bad deal. TSMC has to put lots of money effort and time to maybe turn around the business and completely screw up their capacity LRP. Intel's legacy fabs can't run new processes and TSMC utilization for their legacy processes is subpar with even more fabs coming up in Japan and Germany. So an Intel merger doesn't really save TSMC capex dollars. Intel's EUV fabs are mostly scattered around and while they add up to a good chunk of capacity TSMC'S customers won't accept chips from say Ireland if they did their qualification in Arizona. The biggest benefit to TSMC is killing their most serious competitor since the very early 2000s in the crib before they've had time to prove their capability (or maybe lack thereof).

Since there are already internal and external foundry products designed for 3, 18A and 14A TSMC can't really convert any Intel fabs to TSMC technology for years until the existing production runs start winding dow. So if they really wanted to avoid tariffs they should just accelerate their AZ build out and look at another Greenfield site near Dallas, Boise, Austin, or upstate NY (assuming the local parties are willing to cut a competitive deal my pick would be Dallas given the large TI presence to draw local experience from). That approach should be faster and at the very least cheaper since you aren't paying for any of the assets you don't want.

Also seems like a rotten deal for Intel too since progress seems solid against Intel foundry's financial and technical deliverables (at least based on Intel guidance or any reasonable person's understanding that signs of recovery cannot start to show up until 2026 at the earliest due to long chip design cycles). Why give up after investing tons of cash over the past 3.5 years just before you start generating returns? For two years now Intel said 2024 and especially 2025 would be the years where capex and losses would peak. Any deal TSMC would cut would be a lot smaller than the potential upside, lower Intel's gross margins, and lower their maximum operating margin potential beyond 2026. The market is ripe for sizing with Samsung's repeated failures, TSMC slowing down, and tensions of their leading edge monopoly seemingly rising with TSMC's ever growing cut of the pie (margins).

All in all seems like a lose-lose rather than a win-win.


18A is on the cusp of delivering and possibly with some foundry after. 14A soon after and likely more competitive. The talent at Intel is still the best out there as well has all the support of the vendors. Let’s be honest as big a business TSMC to suppliers and customers everyone wants an alternative. To allow TSMC to managed Intel and transfer technology or manage their fabs will dilute innovation as well as stifle competition.

When you look back beyond EUV TSMC is a follower. Actually from a process architecture not clear where TSMC has innovated.
 
TSMC should look hard at any deals to expand manufacturing rapidly in the US.

For decades, until 2022, Foreign direct investment was all directed toward China, on the assumption of superior future growth. But China's original growth thesis is spent. And the US has a new growth thesis: AI. So investing in the US to get some of the growth, and a better multiple, is something all global CEOs should be considering, TSMC included.
Interesting. This used to be a common story in the Chinese stock market. Some companies would migrate from Nasdaq to the Chinese stock market because their P/E ratios were much higher (3x).
 
Hard to be fired for something you inherited. No I think one of several

1). A whale client said to do a deal he had to go

2). Board wanted to push on a strategy change and it was either get on board or leave.

I am thinking the second as things are pretty dire and require a pivot. They wasted tens of billions on shells that they didn’t need. Foundry should build to real demand not hope and prayer. That is a spectacular failure of Pat’s part also somewhat the board and CFO



18A is on the cusp of delivering and possibly with some foundry after. 14A soon after and likely more competitive. The talent at Intel is still the best out there as well has all the support of the vendors. Let’s be honest as big a business TSMC to suppliers and customers everyone wants an alternative. To allow TSMC to managed Intel and transfer technology or manage their fabs will dilute innovation as well as stifle competition.

When you look back beyond EUV TSMC is a follower. Actually from a process architecture not clear where TSMC has innovated.
its not just architecture Glass Substarte co packaged optics Intel starts the trend and TSMC follows only after Intel unveiled it the rumors on other companies started doing this research.
 
Back
Top