• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Intel Faces Potential Breakup as TSMC and Broadcom Explore Acquisition

Despite all the massive foundry expansion plans, the actual pickup rate is very low and it seems like Intel is still very much the main user of their foundry with the next gen CPU and GPUs rumored to be based on Intel's 18A.
There's already massive capacity expansion across the globe, in fact overcapacity right now & when/if Chinese companies switch (almost) completely inwards you're then sitting on potentially trillions of dollars of dead assets o_O

China, from what I remember, is expanding at a faster rate than Intel/TSMC/Samsung combined although not at the leading edge right now.
 
Well in the grand scheme of capitalism, if your stock/share price drops enough and some person or company has the money in theory they could buy it out...........pending all approvals, etc. first, which in this case would probably be unlikely, especially after that chips "bonus" package.
 
Force whoever buys intel to open up the X86-X64 standard. Monopoly averted.
Speculation of my own, but I think that's not possible unless AMD agrees. Back in the day AMD and Intel came to a cross-licensing agreement and that's why they both use AMD64 and all the CPU extensions and stuff that followed like SSE, FMA, AVX, etc., without hassle or "license warmongering" between them.

A while back someone mentioned that there might be a clause in their agreement with regards to how the licensing would work were Intel or AMD be bought by another party, but that's further speculation since their agreement seems to be private to these days so we apparently don't really know the exact contents.
 
Well in the grand scheme of capitalism, if your stock/share price drops enough and some person or company has the money in theory they could buy it out...........pending all approvals, etc. first, which in this case would probably be unlikely, especially after that chips "bonus" package.
Majority of share holders would need to agree to selling too.

Since for some reason mega behemoths like Blackrock, Vanguard, Berkshire Hathaway, Softbank all own huge stakes in them and each other it's more of a would it make sense to screw around with our properties
 
Well in the grand scheme of capitalism, if your stock/share price drops enough and some person or company has the money in theory they could buy it out...........pending all approvals, etc. first, which in this case would probably be unlikely, especially after that chips "bonus" package.
Some readers here are suspending reality and saying that it is a physical law of the universe that Intel cannot be sold. The real reality is exactly what you said, any publicly traded company can be bought pending approvals. A recent example of that failing is Nvidia trying to buy ARM. A recent example of it succeeding is AMD buying Xilinx. A breakup of Intel where fabs go to TSMC and chip design goes to Broadcom is possible. There are no anti-trust hurdles with this arrangement and as long as TSMC promises that US fabs will be used to make US chips under appropriate security scrutiny, that's not a problem either.

The real hang up around here is change. Intel going away is a big change that is hard for some to accept.
 
Some readers here are suspending reality and saying that it is a physical law of the universe that Intel cannot be sold. The real reality is exactly what you said, any publicly traded company can be bought pending approvals. A recent example of that failing is Nvidia trying to buy ARM. A recent example of it succeeding is AMD buying Xilinx. A breakup of Intel where fabs go to TSMC and chip design goes to Broadcom is possible. There are no anti-trust hurdles with this arrangement and as long as TSMC promises that US fabs will be used to make US chips under appropriate security scrutiny, that's not a problem either.

The real hang up around here is change. Intel going away is a big change that is hard for some to accept.
Huh, Intel is going away?
 
x86 doesn't hold that much value anymore, the third license holder is nowhere to be seen on the market.

Au Contraire, the CLA between AMD and Intel pretty much prevents either from being acquired by a third party without invalidating said agreement. It’s less a value prop and more of just having the ability to continue to exist in the very large desktop and server markets… Even moreso, given that AMD holds the 64 half of x86/64…
 
Huh, Intel is going away?
If they are purchased by another company, then there is no more Intel.

Au Contraire, the CLA between AMD and Intel pretty much prevents either from being acquired by a third party without invalidating said agreement. It’s less a value prop and more of just having the ability to continue to exist in the very large desktop and server markets… Even moreso, given that AMD holds the 64 half of x86/64…
Agreements can be changed or voided.
 
The interesting question in the license agreement topic is: Which company relies more on selling x64 designs, AMD or the potential intel buyer?

And yes, that's a rhetorical one.
 
But seriously, what is the advantage for the market leader in buying part of intel's fabs? *Unless it's to reuse equipment to expand its own production.

Large companies that only design their own chips will never hire Intel to manufacture them. It is impossible to imagine Nvidia or AMD contracting Intel to manufacture chips. If they do this, they would be handing over all their industrial secrets and their release dates for new chips to Intel.
 
Can you imagine what would happen if TSMC did get the Intel fabs and then a certain country finally takes Taiwan back thus acquiring the worlds supermajority of chipmaking capacity.
Yep - that's why I said it's not gonna happen.
 
If they are purchased by another company, then there is no more Intel.
If it's a WD x Sandisk kind of takeover, then as far as the normie will be concerned, Intel would have never left. Intel situation is a bit weird, even with everything happening right now, their CPUs are still everywhere in the consumer space, and the brand image only took a hit in the eyes of tech nerds, but their marketing power is still very strong otherwise. Keeping the name would make things easier...there's not a potential buyer with a strong brand presence with the general public. The new owner killing the Intel name would actually make things easier for AMD :D
 
If it's a WD x Sandisk kind of takeover, then as far as the normie will be concerned, Intel would have never left. Intel situation is a bit weird, even with everything happening right now, their CPUs are still everywhere in the consumer space, and the brand image only took a hit in the eyes of tech nerds, but their marketing power is still very strong otherwise. Keeping the name would make things easier...there's not a potential buyer with a strong brand presence with the general public. The new owner killing the Intel name would actually make things easier for AMD :D
I can see Copilot+ compatible SoCs (big NPU+GPU) making the cut but all other client based products would get axed. Broadcom would probably also keep Gaudi, Altera, Mobileye, networking and edge. But all other data center products would get the axe.

Broadcom might also license x86+Xe to any third party that wants to build AI SoCs.

As for the Intel name, it will hang around long enough for Broadcom to become the new household brand in a similar way as the ATI name lingered for a couple of years.
 
I'm a bit surprised I'm not reading more groaning about the prospect of Broadcom acquiring Intel like when they acquired VMWare.
 
I can see Copilot+ compatible SoCs (big NPU+GPU) making the cut but all other client based products would get axed. Broadcom would probably also keep Gaudi, Altera, Mobileye, networking and edge. But all other data center products would get the axe.

Broadcom might also license x86+Xe to any third party that wants to build AI SoCs.

As for the Intel name, it will hang around long enough for Broadcom to become the new household brand in a similar way as the ATI name lingered for a couple of years.
Core as a standalone CPU is not enough of a dumpster fire to get axed in the near future, Intel still holds over 60% of the markeshare, and that goes up to 70% when it comes to laptop. ARM windows devices are really not that strong (yet) to pose a threat, their performance still rely too much on weither or not you are using an app that's native, and their sales number are still in the single digits.

If core can't become good enough to be a standalone CPU, there's is no way that a core based SOC would be competitive. Especially when the NPU and GPU are still looking weak, and got software issues. Only offering price performance at the low end isn't sutainalbe, ask AMD before the zen era, or any of the other x86 license owner who couldn't keep up. The x86 license is also fairly worthless in the modern era, you also need the x64 license from AMD. Core and ARC must be good enough to stand on their own to have a fighting chance against everything that's coming or is already there. Because AMD and Nvidia x mediatek are not going to pull any punches, and are going to sell SOCs based on strong standalone products. If intel isn't aquired with that mindset, then whoever ends up buying them would have just wasted a few billions for nothing.

AMD already had a strong presence in the conssumer space, but Broadcom ? Broadcom is an obscure company for the general public...Western Digital never put their name on anything related to portable flash storage because they are a computer storage brand, not a Photography/film making brand in the same way that SanDisk has been and still is. Sandisk even got back the PC flash storage.

Edit: what you are saying make sense from the POV of someone who think that Intel should be dismantled for what they represent in term of company/business culture, but as I said before, the general public is largely unaware of that, media don't cover Intel with the same intensity as they do with amazon, meta, Tesla and their work culture/CEOs insanity.

And for the Broadcom name to ever outshine the Intel name, they would have to make a very good chart topper product and make it clear that it happened under their leadership. But mediocre "good enough" releases will never do that. That would be a swansong before they pull out from a market that's too tough for them.
 
Last edited:
I'm a bit surprised I'm not reading more groaning about the prospect of Broadcom acquiring Intel like when they acquired VMWare.
Intel screwed up so badly we just desperately need any competent company to make a better, more competitive go of the GPU and CPU markets.

Core as a standlaone CPU is not enough of a dumpster fire to get axed in the near future, Intel still holds over 60% of the markeshare, and that goes up to 70% when it comes to laptop. ARM windows devices are really not that strong (yet) to pose a threat, their performance still rely too much on weither or not you are using an app that's native, and their sales number are still in the single digits.

If core can't become good enough to be a standalone CPU, there's is no way that a core based SOC would be competitive. Especially when the NPU and GPU are still looking weak, and got software issues. Only offering price performance at the low end isn sutainalbe, ask AMD before the zen era, or any of the other x86 license owner who couldn't keep up. The x86 license is also fairly worthless in the modern era, you also need the x64 license from AMD. Core and ARC must be good enough to stand on their own to have a fighting chance against everything that's coming or is already there. Because AMD and Nvidia x mediatek are not going to pull any punches, and are going to sell SOCs based on strong standalone products. If intel isn't aquired with that mindset, then whoever ends up buying them would have just wasted a few billions for nothing.

AMD already had a strong presence in the conssumer space, but Broadcom ? Broadcom is an obscure company for the general public...Western Digital never put their name on anything related to portable flash storage because they are a computer storage brand, not a Photography/film making brand in the same way that SanDisk has been and still is. Sandisk even got back the PC flash storage.

Edit: what you are saying make sense from the POV of someone who think that Intel should be dismantled for what they represent in term of company/business culture, but as I said before, the general public is largely unaware of that, media don't cover Intel with the same intensity as they do with amazon, meta, Tesla and their work culture/CEOs insanity.

And for the Broadcom name to ever outshine the Intel name, they would have to make a very good chart topper product and make it clear that it happened under their leadership. But mediocre "good enough" releases will never do that. That would be a swansong before they pull out from a market that's too tough for them.
Broadcom has over a trillion dollar market cap value by being 'behind the scenes'. It's the ninth most valued company in the world. Intel has a lot of competing IP in the network, edge and FPGA space that would be valuable to Broadcom. If name recognition is needed, Broadcom can launch a huge ad blitz and even rename the joint venture into Broadtel or something similar.
 
Broadcom might also license x86+Xe to any third party that wants to build AI SoCs.

It would be great for us and for Broadcom.

For us, there would be more x86 CPUs from different brands competing against each other, and thus there would be more competition for performance and lower prices.

And Broadcom would make money from the work of others, just like ARM does.
 
Broadcom has over a trillion dollar market cap value by being 'behind the scenes'. It's the ninth most valued company in the world. Intel has a lot of competing IP in the network, edge and FPGA space that would be valuable to Broadcom. If name recognition is needed, Broadcom can launch a huge ad blitz and even rename the joint venture into Broadtel or something similar.
Yhea, I know that Broadcom is a big company, but for having studied branding and looking at past exemples, that would be a odd choice to do for Broadcom. Rebranding/launching a new B2C brand is a tough exercise, even with money involved... most companies tend to avoid that if possible. There's lot of exemple where a brand name is licensed when a company sold off some of their activities (like Vaio, Nokia...) because the new owner want to profit from the brand power, and if there's a time limit to use the main name of the brand( like Olympus cameras), the new owner is very carefull and try to stay as as close as possible from the older branding to not shake the customers too much. Olympus became OM system, OM being how olympus called their reflex/mirrorless cameras for decades. Even Apple kept the beats brand around.

Broadcom is a very B2B company. When Avago who was also a B2B company bought Broadcom, they also made the decision to keep the Broadcom name because that name got more weight to it. Unless Broadcom want to become know as a B2C company, that would be a surprising decision to make, having to build consumer momentum from scratch, and going up agaisnt more estasblished very competent brands.

Consumers are tricky, if they had positives impressions associated to a brand, they tend to not react very well to change that feels too drastic. Tons of uncertainty, when I worked in an advertising agency, it wasn't that rare to see a customer who wanted a rebrand but was scared shitless about how the public would react to it. They tend to favor things that feels like an evolution rather than a transformation. Unless the brand image was so weak/negative to begin with, changing it was actually mandatory to save the business.
 
Intel screwed up so badly we just desperately need any competent company to make a better, more competitive go of the GPU and CPU markets.
Well said, but you're quite a joker here. Since basically anyone would be better than Intel itself at designing chips at this point – They just lost their slack and any expertise well over a decade ago.

Broadcom would eventually help correcting the course towards any sane return to normalcy, save what salvageable and rescue given division, which ain't too much corrupted to be saved yet – Tossing the rest for cents on a dollar afterwards and help to smoke out Intel's decades-old well-trotted paths of everlasting complacency and hubris.

Since for a start, I can't really remember too much nonsense, broken devices or sh!tty implementations from Broadcom they made yet. Unlike Intel, Broadcom hasn't single-handedly bricked millions of mainboards with their flawed Broadcom-NICs nor did they turned tens of millions of Raspberry Pies all around the globe into basically eWaste with their flawed chips, like Intel does on the regular!

Broadcom has over a trillion dollar market cap value by being 'behind the scenes'. It's the ninth most valued company in the world.
Didn't we all knew that Broadcom was prone to act upon it and that Trillion-dollar valuation and wouldn't've had let that chance of a life-time passing by unused, as their valuation rose towards 1Tr USD?
Intel has a lot of competing IP in the network, edge and FPGA space that would be valuable to Broadcom.
I think the only useful for Broadcom would be especially everything network and maybe edge.
FPGAs I think could be already a tad bit off for Broadcom and possibly gets sold anyway to Lattice, MicroChip or Achronix.
If name recognition is needed, Broadcom can launch a huge ad blitz and even rename the joint venture into Broadtel or something similar.
No offense here, but that's the for sure the dumbest thing I've had to read the whole week now! You're joking, right?

Their own Broadcom-brand is times more valuable and has a way higher actual brand-recognition than anything Intel by now, especially in the actually professional field of networking and computing. Meanwhile Intel has burned most of their actual reputation the last years, notably since AMD's Ryzen, with brain-dead, mud-throwing and back-handed false marketing …

Also don't underestimate the huge societal changes society went through the recent years via gaming and everything technology!
The societal awareness before technology in general and especially anything computing has newer been so high and at the forefront of society – Everyone now knows what a computer is and how Office works! This isn't the 80s or 90s anymore, were only geeks, computer hobbyists or programmer were aware of computers and what a processor is. The last two generations being raised using consoles and today's kids battling each other in chat over their consoles' specs is no joke to play with. People are more informed than ever before.

Since most actually informed people already well-deservedly associate anything Intel since 2018 either with hardware, which is riddled with never-ending security-flaws or a low-performing expensive wannabe-deluxe alternative as a neat remembrance of the past – Even at consumers, the reputation is down to a all-time low. Their CPUs literally dying now is also something which will stick in the back of the head for a long, long time at end-suers nor will the influential crowd of informed end-users forget Intel-CPUs as the prominent power-hungry heating-devices with embedded calculator-functions for years to come.

In the corporate world, at enterprises and other professional businesses, Intel now has to look upon a almost foreign territory of scorched earth for them. No-one really trusts them anymore nor will the corporate world of professionals forget anytime soon, how Intel jacked up their price-tags during the completely self-inflicted shortages over the aftermath of Meltdown, Spectre, Foreshadow & Co. The overtime and stress Intel brought over the computing world, when their flawed CPUs had to be basically halved in their compute-capabilities world-wide overnight (over Intel's utterly broken Hyper-Threading), won't be forgotten that quick, let me tell you that.

Worse, even the crowd of public pencil-pushers, usually deaf to any technology, as well as the corporate generally and proudly tech-unsavvy yet sharp-minded bean-counters, had to take suddenly notice of Intel's jacked up price-tags, when the crowds of paper-pushers from accounting all of a sudden had to re-schedule bugdets (at overtime, minutes past regular working hours!!), which were otherwise since decades set once at the start of the fiscal period of the department.

No … Today, if anything, Intel is mostly known for "quality and performance, like it used to be" only still around the circles of (often bribed) academics (or other highly-educated rented loudmouths), and their even less informed corporate counterparts of perfect idiots (or the usual crowd of not even remotely any tech-savvy complete morons of most public IT-departments).

Other than that, especially Intel's Xeon-brand in the server-space, is already basically mostly dead and a reason to look for alternatives – VMware's utter price-jacking also didn't really helped to leave the argument of any strategy of platform-homogeneity as a final kicker for Intel. Never mind Intel's inferior metrics for years on end on price, performance, efficiency and power-draw, cooling-ability coolability, security or the corporate's world and businesses' need for predictability, with Intel having ever so often sneakily evovled their road-maps into factual "rolling releases".


That being said, if that deal goes through with Broadcom eventually taking over the whole of Intel's former product group and basically Intel as a whole as is, save anything manufacturing … Then I wouldn't be even remotely surprised, if Broadcom not soon after ditches everything datacenter-CPU aka Xeon as a whole to either IBM (if they even feel like it, for reasons of compatibility), Google (to play with it) or Hewlett Packard Enterprise (HPE, for nostalgic reasons of rememberance).

… or in a brilliant move of royal cajolery and rip-off to Facebook for a unheard of sum, while they somehow managed to convince Facebook, that they now would "never again have to buy any CPUs ever again"!

Since not only is the Xeon brand as a whole exremely tranished by now in the business-world, but even the technical base for it from Intel is way too outdated technologically, to keep it running and bring forward any new products, to keep it within the portfolio – Save the few low-volume halo-SKUs, which are expensive asf to manufacture anyway.

Thus it would be basically mostly dead capital for Broadcom, so the chance of them selling Xeon as a whole I would consider as not really low.
 
Can't Intel just sell the factories to an American investment group and return to the government the money it received from the "CHIPS and Science Act"?
 
Can't Intel just sell the factories to an American investment group and return to the government the money it received from the "CHIPS and Science Act"?
I doubt they received any of that money, they probably went to the bank with that promise to get a loan until they get it.
 
I doubt they received any of that money, they probably went to the bank with that promise to get a loan until they get it.
Intel per their earning releases in January notified the SEC, that they have gotten $1.1Bn in Dezember 2024 and again $1.1Bn this January.
"Dave Zinsner, Intel’s co-interim CEO, executive vice president, and CFO, said the Silicon Valley-based company received the first tranche of $1.1 billion in federal grants at the end of 2024 and an additional $1.1 billion in January 2025." — TechCrunch · Intel has already received $2.2B in federal grants for chip production

So, knowing Intel's well-known Financial engineering™ they're notoriously known and infamous for ever since, from these $14.3Bn posted as revenue in 4Q24 this January, you can already deduct $2.2Bn … What Intel also artificially upped their revenue with, what that Intel already received +$500m in November last year from the EU.
"U.S. chipmaker Intel (INTC.O), opens new tab has received 515.55 million euros ($536 million) in default interest from EU antitrust regulators related to a 1.06 billion euro fine levied in 2009, EU antitrust chief Teresa Ribera has told EU lawmakers." — Reuters · Intel gets $536 million in interest from 1.06 billion euro EU antitrust fine

To no greater surprise Intel intentionally did not disclose that during their earnings call, to prop the balance-sheet and avoid nasty questions…
Everyone can imagine my utter shock about Intel's appalling dishonesty here. I was shocked. Shocked!
 
Back
Top