Intel's $13B foundry loss raises stakes for potential TSMC partnership

[Skye Jacobs]

In context: Once the undisputed leader in semiconductor manufacturing, Intel now finds itself at a critical juncture as its foundry operations face significant financial challenges. It remains uncertain whether a deal with TSMC can rescue Intel's foundry business, but without it the company – better known in its heyday as "Chipzilla" – must find a way to address its manufacturing challenges and financial losses.

Intel's foundry division reported a staggering loss of over $13 billion on $17.5 billion in revenue last year. In Q2 2024 alone, the foundry posted an operating loss of $2.83 billion, a sharp increase from $1.87 billion the previous year. This stands in stark contrast to TSMC, the industry leader, which generated $41.1 billion in operating profit on $90 billion in revenue during the same period. These figures highlight the severity of Intel's predicament.

As Intel grapples with its foundry woes, speculation has emerged about a potential partnership with TSMC. This idea gained traction following a report from Robert W. Baird analysts, citing "discussions from the Asia supply chain," which suggested that TSMC could become a joint owner of Intel's manufacturing business after a potential spinoff. While unconfirmed, the possibility of such a collaboration has drawn significant interest from industry observers.

Chris Caso, an analyst at Wolfe Research, spells out the rationale behind this potential partnership: Intel's core server and PC businesses will no longer generate enough growth to absorb the significant costs for leading-edge fabs, he said in a research note. Caso further emphasizes that only TSMC can drive the foundry volume needed to absorb Intel's fixed costs in an expeditious manner.

At the same time, Intel's financial struggles have taken a toll on its market value. The company's stock price plummeted 60% last year, recently trading near a 10-year low. Even with a recent 22% surge in share price, Intel's market capitalization remains roughly one-eighth that of TSMC's – a stark reversal from just five years ago when both companies were valued at parity.

Adding to Intel's woes is its substantial cash burn. Over the past three years, the company has spent nearly $40 billion in an effort to catch up with TSMC's manufacturing processes. According to FactSet estimates, analysts expect negative free cash flow to persist through at least the end of next year.

Beyond financial concerns, Intel's foundry operations also suffer from a technological lag. The company trails TSMC by approximately a year in achieving competitive yields for each new process node. Additionally, Intel's manufacturing costs are estimated to be 30% to 35% higher than TSMC's due to lower wafer volumes.

A lot hinges on Intel's latest 18A manufacturing process, which is expected to be a pivotal moment for the company's foundry ambitions. Intel has positioned 18A as a game-changer, boasting advancements such as RibbonFET transistors and PowerVia technology to enhance power efficiency and performance.

Permalink to story:

Intel's $13B foundry loss raises stakes for potential TSMC partnership

 

[yRaz]

I've been managing my own portfolio for almost 20 years now and I've noticed 2 very similar trends. One is that when people talk about how safe an investment is and shill for it constantly, it's usually close to a crash. The other is that when otherwise successful companies start making the news for how bad they're doing, it's usually because someone wants to jump in and buy it for pennies on the dollar.

While Intel is most certainly having it's problems, it's not going to go away anytime soon.

nVidia has had so many cancelations for blackwell from AI hyperscalers they are having a difficult time using all the capacity they bought from TSMC. We ARE going to see a slowdown in demand for cutting edge nodes from TSMC, Intel might be able to by some of that excess capacity.

The thing is, Intel is both a Chip design and a Fab. Intel doesn't need to be the best to turn things around, they just have to make money. Short term, they make their desktop chips with TSMC while working on their nodes. And here is the thing about the nodes, they can make features BIGGER than what they're rated to do, not smaller. Intel can still make 10-14NM chips on their cutting edge equipment with very high yields to ASICs or things like RapsPis where power efficency and being the fastest don't matter while also dialing in their processes. 2% increase here, 5% here, half a percent there. Those all add up and that's what TSMC has been doing for the last 10 years, adding up every tiny percent.

Intel is capable of that and has many talented people. Now that Pat is gone, maybe they can start focusing on making chips again instead of marketing. I've never seen a TSMC advertisement in my life, but I have been seeing "intel inside" advertisements my entire life. If Intel had spent that money on trying to be the best instead of on brand recognition, well, AMD might not be here and nVidia would never shove us into the AI Rectum we are forced to smell on a daily basis.

 

[daffy duck]

This would be a disaster for consumers and TSMC would be the last one that I would allow to own Intel Fabs.

 

[user556]

nVidia has had so many cancelations for blackwell from AI hyperscalers they are having a difficult time using all the capacity they bought from TSMC. We ARE going to see a slowdown in demand for cutting edge nodes from TSMC, Intel might be able to by some of that excess capacity.

That's not going to be a great help to Intel's idle money sinks though.

The thing is, Intel is both a Chip design and a Fab. Intel doesn't need to be the best to turn things around, they just have to make money. ...

Funnily, if it wasn't for AMD, Intel could have done exactly that. Carried on making money off the captive PC market irrespective of other advancements outside of the PC.

... I've never seen a TSMC advertisement in my life, but I have been seeing "intel inside" advertisements my entire life. If Intel had spent that money on trying to be the best instead of on brand recognition, well, AMD might not be here and nVidia would never shove us into the AI Rectum we are forced to smell on a daily basis.

Intel of course has partaken in plenty of anticompetitive behaviours but AMD had to stay or Intel would have been dragged over the anticompetitive coals back then. Different story now of course.

 

[ruid84]

Intel can still make 10-14NM chips on their cutting edge equipment with very high yields to ASICs or things like RapsPis where power efficency and being the fastest don't matter while also dialing in their processes. 2% increase here, 5% here, half a percent there. Those all add up and that's what TSMC has been doing for the last 10 years, adding up every tiny percent.

The issue here is that TSMC is able to offer better value also in the lower end segment, where the economy of scale is even more important(due to low margins) compared to the high end. TSMC is even trying to outcompete GlobalFoundries.... Therefore, Intel is pretty much forced to develop a cutting edge node and and offer it at a premium price.

 

[user556]

Therefore, Intel is pretty much forced to develop a cutting edge node and and offer it at a premium price.

Only because AMD is using TSMC.

 

[Bamda]

I've been managing my own portfolio for almost 20 years now and I've noticed 2 very similar trends. One is that when people talk about how safe an investment is and shill for it constantly, it's usually close to a crash. The other is that when otherwise successful companies start making the news for how bad they're doing, it's usually because someone wants to jump in and buy it for pennies on the dollar.

While Intel is most certainly having it's problems, it's not going to go away anytime soon.

nVidia has had so many cancelations for blackwell from AI hyperscalers they are having a difficult time using all the capacity they bought from TSMC. We ARE going to see a slowdown in demand for cutting edge nodes from TSMC, Intel might be able to by some of that excess capacity.

The thing is, Intel is both a Chip design and a Fab. Intel doesn't need to be the best to turn things around, they just have to make money. Short term, they make their desktop chips with TSMC while working on their nodes. And here is the thing about the nodes, they can make features BIGGER than what they're rated to do, not smaller. Intel can still make 10-14NM chips on their cutting edge equipment with very high yields to ASICs or things like RapsPis where power efficency and being the fastest don't matter while also dialing in their processes. 2% increase here, 5% here, half a percent there. Those all add up and that's what TSMC has been doing for the last 10 years, adding up every tiny percent.

Intel is capable of that and has many talented people. Now that Pat is gone, maybe they can start focusing on making chips again instead of marketing. I've never seen a TSMC advertisement in my life, but I have been seeing "intel inside" advertisements my entire life. If Intel had spent that money on trying to be the best instead of on brand recognition, well, AMD might not be here and nVidia would never shove us into the AI Rectum we are forced to smell on a daily basis.

Maybe you should apply for the CEO position. Pat thought it was going to be easy to put AMD in the rear-view mirror. LOL

 

[yRaz]

Maybe you should apply for the CEO position. Pat thought it was going to be easy to put AMD in the rear-view mirror. LOL

Pat made the same mistake IBM made with mainframes and ignoring the PC. Pat thought the cloud was the future and all computing was going to be done in the cloud. Intel always did a good job on the mobile market and their mobile chips are still better than AMDs. Pat focused on the cloud and ended up, ironically, losing their server business to AMDs desktop chips when they were turned into chiplets and scaled up.

Now we're seeing that people are sick of the cloud and AI. Pat, while not a fool, believed the future was the cloud. He bet the company on it and lost.

I cashed out the last of my NV stock and opened a fairly large position on Intel. I 20X'd my money on NV, I don't care if it goes up more. I will never feel bad about taking profit at 2000%. But Intel and a few other companies are peaking my interest lately and I also feel most of them are "at the bottom" so I'm not exactly worried about losing money in a crash right now.

And while not a financial advisor, I'm looking at all the intel talk and thinking about who Intel actually is as a company, it's a bargain right now. Atleast to me. If they are still doing bad in a year or two then I'll change my opinion. The thing is, I don't think they'll be doing bad in a few years. And even if they goto zero, I still have that other 1900% I made on NV

 

[gamerk2]

I cashed out the last of my NV stock and opened a fairly large position on Intel. I 20X'd my money on NV, I don't care if it goes up more. I will never feel bad about taking profit at 2000%. But Intel and a few other companies are peaking my interest lately and I also feel most of them are "at the bottom" so I'm not exactly worried about losing money in a crash right now.

I brought AMD back when it was $1.68. I thought I was a genius selling at $15, lol.

But yes, find good companies near the bottom of their value; either they recover and you make bank, or they get absorbed and you get your money back plus a small premium. Not like a company like Intel will outright go under; they'd get nationalized before that ever happens.

 

[yRaz]

I brought AMD back when it was $1.68. I thought I was a genius selling at $15, lol.

But yes, find good companies near the bottom of their value; either they recover and you make bank, or they get absorbed and you get your money back plus a small premium. Not like a company like Intel will outright go under; they'd get nationalized before that ever happens.

I'm just looking at the market and already saw that NV was already extremely overvalued and then I saw all the cancelations for Blackwell and decided it was a good time to get out. My NV position had outdone the rest of my portfolio in the last ~7 years than my portfolio has done combined in the last 20 years.

AMD has never been in the top 5 most valuable companies before. Aside from Apple, I don't see any of the top 5 holding their value for very long.

And let's also not forget AMDs mistake of getting comfortable in their position before Conroe/core2duo era. AMD made the same mistake then that Intel is making now. I see NV Blackwell as a warning sign of a repeat of what AMD did then and what Intel did recently. The difference is that I can see a redemption arc with Intel now that Pat is gone

 

[gamerk2]

I'm just looking at the market and already saw that NV was already extremely overvalued and then I saw all the cancelations for Blackwell and decided it was a good time to get out. My NV position had outdone the rest of my portfolio in the last ~7 years than my portfolio has done combined in the last 20 years.

AMD has never been in the top 5 most valuable companies before. Aside from Apple, I don't see any of the top 5 holding their value for very long.

And let's also not forget AMDs mistake of getting comfortable in their position before Conroe/core2duo era. AMD made the same mistake then that Intel is making now. I see NV Blackwell as a warning sign of a repeat of what AMD did then and what Intel did recently. The difference is that I can see a redemption arc with Intel now that Pat is gone

I remember the Phenom argument: "Core 2 Quad is just two Core 2's stitched together; Phenom is a TRUE Quad Core and therefore better!".

But yeah, between the E8600 and Q6600, AMD was basically shut out for an entire generation.

 

[Bamda]

Pat made the same mistake IBM made with mainframes and ignoring the PC. Pat thought the cloud was the future and all computing was going to be done in the cloud. Intel always did a good job on the mobile market and their mobile chips are still better than AMDs. Pat focused on the cloud and ended up, ironically, losing their server business to AMDs desktop chips when they were turned into chiplets and scaled up.

Now we're seeing that people are sick of the cloud and AI. Pat, while not a fool, believed the future was the cloud. He bet the company on it and lost.

I cashed out the last of my NV stock and opened a fairly large position on Intel. I 20X'd my money on NV, I don't care if it goes up more. I will never feel bad about taking profit at 2000%. But Intel and a few other companies are peaking my interest lately and I also feel most of them are "at the bottom" so I'm not exactly worried about losing money in a crash right now.

And while not a financial advisor, I'm looking at all the intel talk and thinking about who Intel actually is as a company, it's a bargain right now. Atleast to me. If they are still doing bad in a year or two then I'll change my opinion. The thing is, I don't think they'll be doing bad in a few years. And even if they goto zero, I still have that other 1900% I made on NV

Like I said, the CEO position is still open. Go apply now!

 

[enemys]

Pat made the same mistake IBM made with mainframes and ignoring the PC. Pat thought the cloud was the future and all computing was going to be done in the cloud. Intel always did a good job on the mobile market and their mobile chips are still better than AMDs. Pat focused on the cloud and ended up, ironically, losing their server business to AMDs desktop chips when they were turned into chiplets and scaled up.

Now we're seeing that people are sick of the cloud and AI. Pat, while not a fool, believed the future was the cloud. He bet the company on it and lost.

I cashed out the last of my NV stock and opened a fairly large position on Intel. I 20X'd my money on NV, I don't care if it goes up more. I will never feel bad about taking profit at 2000%. But Intel and a few other companies are peaking my interest lately and I also feel most of them are "at the bottom" so I'm not exactly worried about losing money in a crash right now.

And while not a financial advisor, I'm looking at all the intel talk and thinking about who Intel actually is as a company, it's a bargain right now. Atleast to me. If they are still doing bad in a year or two then I'll change my opinion. The thing is, I don't think they'll be doing bad in a few years. And even if they goto zero, I still have that other 1900% I made on NV

Wow you're so cool and so wealthy, but can you please keep quiet while people talk about actual semiconductors, not your financial gains?

 

[yRaz]

Wow you're so cool and so wealthy, but can you please keep quiet while people talk about actual semiconductors, not your financial gains?

oh, jeez, I was just about to show everyone the diamond encrusted buttplug using b200 chip I had made using vapor deposition. Sorry about that, I had trouble reading the room