News Intel CEO Pat Gelsinger retires, effective immediately — also steps down from BOD, two co-CEOs step in

[JarredWaltonGPU]

Intel's issue was/is process technology, Apple absorbed chip design engineers and it has nothing to do with process. Old white men targeted layoff is completely false.


Apple did/does not charge premium! Apple makes money because of its scale, if you want to buy an Android phone with same/similar quality, it costs more.

Phones used to cost $200 or less, before the smartphone era. So yeah, that's definitely premium pricing. And again, that doesn't mean other brands don't try to do the same thing (Samsung being a prime example).

This has nothing to do with ASML developing high resolution EUV machines and TSMC then building advanced process nodes based on those machines, like nothing at all.

Apple is just a luxury fashion designer, a very good one but still just that. And the Apple ARM SOC isn't anything special, no matter how much people were paid to rave about it. Apple paid TSMC a ton of money to be the first customers on any new production node, meaning their Mx CPU's were a generation ahead of AMD and several generations ahead of Intel.

As I said before, people are doing mental gymnastics to make their favorite luxury brand into something they aren't. Unless Apple invented a time machine, they were not involved in the aforementioned R&D. Of course I wouldn't put it past people to insist Steve Jobs did create such a thing.

As for Intel's recent financial struggles, it's because years ago they missed the boat on incorporating EUV Into their own fabrication process. The yields on their 10nm process were so bad they had to stack with the 14nm process and just keep releasing products on that while AMD and Apple both got on TSMC's 7nm process. If a CPU on 7nm can't absolutely crush one on 14nm then those designers need to be fired. And even though Intel has now moved past those issues, they are still an entire generation or more behind TSMC in fabrication technology. This means their competitors, AMD / Apple, will always be at least a generation ahead.

That is why Intel is thinking of doing what AMD did and spinning off it's fabrication business into a separate entity. Most microprocessors are not these high performance components that need to be on the latest generation process to be competitive. Intel's vertical integration is what's making it uncompetitive in the processor space. There are places that works but this isn't one of them.

There are tons of factors in Intel's struggles. It's not just process nodes and EUV. That's a part of it for sure, a big part. It's not just a lack of GPU and AI success, but that's another big part. When any company disrupts the market — as we've seen with smartphones, EUV, GPUs, AI, chiplets — it impacts others.

You talk about doing mental gymnastics to say that Apple has played a role, and yet you're going the opposite direction and doing gymnastics to pretend it had no effect. I'm not even saying Apple was a major role or the primary role. It's simply one part of many things going wrong.

It seems you dismiss Apple because you don't like them. I dislike Apple as well for a variety of reasons, the only Apple product I have is a company provided iPhone. But I can still see clearly that as a major tech company, one of the most valuable in the world, Apple has had an impact on Intel. It's blind to pretend otherwise.

And while we're at it, Google / Alphabet has had an impact on Intel as well. Big or small, it doesn't really matter. Any impact at all counts. Nvidia has certainly had an impact. AMD, though far smaller than those other three I've named, has had an impact. The biggest tech companies all play a role in how other tech companies are doing. eg, Is Google doing more or less business with Intel than in the past? If it's less, that's impactful.

If Intel sells off it's foundry business, who will step in to make that foundry business successful? Basically, the only solution I see there is government intervention, forcing the use of Intel foundries and putting billions into improving those foundries. Because the foundries alone cost billions to create, and they need to be run at full capacity to pay back on the investment.

 

[palladin9479]

Phones used to cost $200 or less, before the smartphone era. So yeah, that's definitely premium pricing. And again, that doesn't mean other brands don't try to do the same thing (Samsung being a prime example).

There are tons of factors in Intel's struggles. It's not just process nodes and EUV. That's a part of it for sure, a big part. It's not just a lack of GPU and AI success, but that's another big part. When any company disrupts the market — as we've seen with smartphones, EUV, GPUs, AI, chiplets — it impacts others.

You talk about doing mental gymnastics to say that Apple has played a role, and yet you're going the opposite direction and doing gymnastics to pretend it had no effect. I'm not even saying Apple was a major role or the primary role. It's simply one part of many things going wrong.

It seems you dismiss Apple because you don't like them. I dislike Apple as well for a variety of reasons, the only Apple product I have is a company provided iPhone. But I can still see clearly that as a major tech company, one of the most valuable in the world, Apple has had an impact on Intel. It's blind to pretend otherwise.

And while we're at it, Google / Alphabet has had an impact on Intel as well. Big or small, it doesn't really matter. Any impact at all counts. Nvidia has certainly had an impact. AMD, though far smaller than those other three I've named, has had an impact. The biggest tech companies all play a role in how other tech companies are doing. eg, Is Google doing more or less business with Intel than in the past? If it's less, that's impactful.

This is just six degree's to Kevin Bacon.

https://en.wikipedia.org/wiki/Six_Degrees_of_Kevin_Bacon

Using such large amorphous definitions allows anyone to say anything about anything else and have it not be wrong. That logic leads to a storm in the Atlantic being responsible for Intel's woes and a can discarded by an Apple employee being the cause of that storm.

I refute such open ended "everything is responsible for everything else" and stick to first principals and direct causation. Intel is in a large financial pickle because they can no longer use stock to generate capital to grow and invest. They are in that position because they can not generate enough revenue to cover the massive expenses of their foundry. This means the financial issues are a directly related to running an unprofitable foundry and not selling enough inventory, end of story.

AMD is doing fine selling inventory so it's not "market forces BS" like tablets / smartphones / AI-smokescreens. And it's certainly has absolutely nothing to do with Apple. TSMC is the largest foundry on the planet and is doing extremely well with sales so it's not a consumer market slump.

As for Apple, I can appreciate them as a very market savvy and capable company. They discovered the secret sauce to sell commodity hardware at large markups by using branding. This is the same business model that Chanel, Louis Vuitton, Versace, and Bottega use. Apple consumers on the other hand tend to have this burning desire to justify buying into that designer brand. This is most easily demonstrated by simply asking them how they using their Apple PC.

If Intel sells off it's foundry business, who will step in to make that foundry business successful? Basically, the only solution I see there is government intervention, forcing the use of Intel foundries and putting billions into improving those foundries. Because the foundries alone cost billions to create, and they need to be run at full capacity to pay back on the investment.

As I mentioned above the vast majority of integrated circuits on the planet are not using the high resolution / high density fabrication process's. There is no need for the processor in your watch, refrigerator or TV to be at 7nm and smaller. As a PC site you are largely focused on the next generation, most recent released super duper consumer gaming enthusiast product, and those are in the minority of all consumer products that get shipped. When AMD spun off GloFo, that entity didn't just vanish, they kept going as a separate entity that could not only do CPU's but now was free to produce everything else.

ASML's low-NA EUV system used for last generation microprocessors (7nm) costs about $180 million each. Their recent high-NA EUV is required for 3nm microprocessors and costs $380 million each. The next generation Hyper-NA EUV, which should be available around 2030 and is required for sub 1nm nodes, is going to cost around $760~800 million each. Making equipment for high performance processors is ridiculously expensive, this is why all the other foundries have lagged and we just have TSMC and Samsung at the top end and even then Samsung lags behind and is very careful about their markets.

After being spun off GloFo struggled but then found it's footing and is doing well, Intel Foundry will be fine too.

https://www.macrotrends.net/stocks/charts/GFS/globalfoundries/gross-profit

https://investors.gf.com/news-relea...-reports-third-quarter-2024-financial-results

 

[thestryker]

Their recent high-NA EUV is required for 3nm microprocessors and costs $380 million each.

Nobody is using High-NA for "3nm" nodes and TSMC isn't planning on using it until after A16.

After being spun off GloFo struggled but then found it's footing and is doing well, Intel Foundry will be fine too.

I'm not sure why you think looking at GloFo is a positive. Leading edge drives technology forward and losing them as a competitor hasn't helped the market. Losing Intel as a leading edge competitor would realistically just leave TSMC and I'm not sure why any rational person would think this is a good outcome. Samsung cannot actually compete with TSMC as despite their GAA nodes taping out earlier they've not been used in any products that can justify their existence because yields are terrible. I sure wouldn't be happy with Intel being relegated to just another generic foundry and nobody else should be either.

 

[palladin9479]

Nobody is using High-NA for "3nm" nodes and TSMC isn't planning on using it until after A16.

Ehh people absolutely are and that includes TSMC.

[News] Price for ASML’s Hyper-NA EUV Rumored to Double, Causing TSMC, Samsung and Intel to Hesitate | TrendForce News

Semiconductor equipment giant ASML plans to launch Hyper-NA Extreme Ultraviolet (EUV) machines by 2030, signaling the advent of the Angstrom era for s...

www.trendforce.com

Intel was the first foundry to adopt high-NA EUV technology. Last year, its foundry business suffered a USD 7 billion loss, and in the first quarter of this year, it faced a record operational loss. One of the reasons for these financial challenges may be contributed to the cost burden of being an early adopter of the next-generation EUV equipment.

ASML has stated that high-NA EUV will enable Intel to produce chips with process nodes from 2 nanometers down to 14 angstroms (1.4 nanometers) and from 10 angstroms (1 nanometer) down to 7 angstroms (0.7 nanometers). ASML also mentioned that Hyper-NA will be essential for future angstrom-scale processes, as it can reduce the risks associated with multi-patterning processes, the report noted.

https://www.tomshardware.com/tech-i...ion-this-month-according-to-industry-insiders

Samsung has opted to skip High-NA and go to Hyper-NA when it becomes available.

TSMC's plan is to only use High-NA for extreme performance parts (Apple, nVidia, AMD, maybe Intel) and stay with the older equipment for non-performance parts, which as I've mentioned is the majority of what gets made.

As for your statement about Intel, their foundry may or may not be able to compete with TSMC and Samsung, they definitely can not compete if they can only produce for Intel. You, and others, seem to be on this mindset that if your not first your last, which is not how markets work. Doing all the low performance mass production parts is what provides stable income and enables a foundry to invest into next generation fabrication techniques.

 

[JarredWaltonGPU]

This is just six degree's to Kevin Bacon.

https://en.wikipedia.org/wiki/Six_Degrees_of_Kevin_Bacon

Using such large amorphous definitions allows anyone to say anything about anything else and have it not be wrong. That logic leads to a storm in the Atlantic being responsible for Intel's woes and a can discarded by an Apple employee being the cause of that storm.

I refute such open ended "everything is responsible for everything else" and stick to first principals and direct causation. Intel is in a large financial pickle because they can no longer use stock to generate capital to grow and invest. They are in that position because they can not generate enough revenue to cover the massive expenses of their foundry. This means the financial issues are a directly related to running an unprofitable foundry and not selling enough inventory, end of story.

AMD is doing fine selling inventory so it's not "market forces BS" like tablets / smartphones / AI-smokescreens. And it's certainly has absolutely nothing to do with Apple. TSMC is the largest foundry on the planet and is doing extremely well with sales so it's not a consumer market slump.

Okay, I'm not saying *any and every* impact. Thanks for putting up the strawman. You've seen my writing enough to know I'm not an idiot I hope. That's why I feel you're intentionally being obtuse here and refusing to admit that there are lots of factors in Intel's problems.

A 0.01% hit to the bottom line doesn't matter, I'm sure we can all agree. If you think Apple's business to Intel when it was using x86 wasn't at all meaningful, though? There were chipsets and processors involved. Apple sells something like 4~6 million MacBooks a year, sometimes more. That would probably be in the range of $100 to $200 compensation for Intel per laptop, possible even more. They have been the "most popular single laptop model" for a long time now (because Apple only does like three models instead of the 50 or 100 that Dell, HP, etc. might have.) That. Matters.

A billion dollars or more in profits for a company, even one as large as Intel, still matters. That's a non-negligible amount. (If it were a trillion dollar company, then I'd say it didn't matter, but I'm not sure Intel ever did much more than $100 billion in revenue a year, so losing 1% or more of that can and will matter if it's not replaced. There's also the hit to the reputation. If Apple doesn't think Intel is worth using, there are a bunch of less tech savvy people that will believe Apple was right, whether it was or not. How much of a hit to the brand was Apple leaving, in light of that? Again, I'd postulate that it was absolutely not "nothing at all." It was something, probably several somethings.

And it's not just foundries creating the problem, as AMD and Nvidia prove. If Intel's foundries were the only issue, then chips produced elsewhere would be doing much better. But Alchemist, while a decent first attempt for dedicated GPU, was not a massive success — or even a success at all, based on what I've seen in the market. Raptor Lake sold well but has plenty of problems, particularly over the long haul. Arrow Lake probably won't be a major success, even though it's primarily TSMC produced on the cutting edge N3B node. And the difficulties all of those face were not purely constrained to the process nodes used. There were design issues and concerns with each of those mentioned product lines.

The foundries are a big drain on finances when your products aren't doing well. No one disputes that. But trying to distill this down to just *one* thing — like EUV or the lack thereof — that caused Intel to miss the boat is naive in the extreme. Everything compounds. Even if Apple only had a 1% impact, and Google had a 1% impact, and AMD, Nvidia, Qualcomm, tablets, smartphones, AI, chiplets, etc. each have a 1% impact on Intel's bottom line, compound that all together and you can easily get a 10% change or more.

Was EUV alone a 10% change for Intel, in the negative direction? Almost certainly, probably a lot more than that over time. But Intel was also trying to lead from the front, and when it lost its way at 14nm and struggled for five years, that had far reaching rammifications. The company lost the process node leadership after 14nm, where it was two years ahead, to 10nm, where it was two years behind. And that wasn't purely about EUV, it was about all the other stuff going wrong with 10nm as well.

Which is all to say, getting back to the subject at hand, that Intel made massive mistakes in the past decade. Pat Gelsinger wasn't perfect either, but if Intel was ever going to regain the lead, it needed to make massive changes. I'm not sure it's even going to keep trying to regain the lead without Gelsinger, but we'll see. Selling off the foundries as a separate entity could work out in the long run for Intel, but I'm not so sure that will be good for the US, given the current geopolitical situation. But hey, we've got that amazingly successful TSMC Arizona plant that will be great, right? LOL

 

[palladin9479]

Okay, I'm not saying *any and every* impact. Thanks for putting up the strawman. You've seen my writing enough to know I'm not an idiot I hope. That's why I feel you're intentionally being obtuse here and refusing to admit that there are lots of factors in Intel's problems.

A 0.01% hit to the bottom line doesn't matter, I'm sure we can all agree. If you think Apple's business to Intel when it was using x86 wasn't at all meaningful, though? There were chipsets and processors involved. Apple sells something like 4~6 million MacBooks a year, sometimes more. That would probably be in the range of $100 to $200 compensation for Intel per laptop, possible even more. They have been the "most popular single laptop model" for a long time now (because Apple only does like three models instead of the 50 or 100 that Dell, HP, etc. might have.) That. Matters.

A billion dollars or more in profits for a company, even one as large as Intel, still matters. That's a non-negligible amount. (If it were a trillion dollar company, then I'd say it didn't matter, but I'm not sure Intel ever did much more than $100 billion in revenue a year, so losing 1% or more of that can and will matter if it's not replaced. There's also the hit to the reputation. If Apple doesn't think Intel is worth using, there are a bunch of less tech savvy people that will believe Apple was right, whether it was or not. How much of a hit to the brand was Apple leaving, in light of that? Again, I'd postulate that it was absolutely not "nothing at all." It was something, probably several somethings.

And it's not just foundries creating the problem, as AMD and Nvidia prove. If Intel's foundries were the only issue, then chips produced elsewhere would be doing much better. But Alchemist, while a decent first attempt for dedicated GPU, was not a massive success — or even a success at all, based on what I've seen in the market. Raptor Lake sold well but has plenty of problems, particularly over the long haul. Arrow Lake probably won't be a major success, even though it's primarily TSMC produced on the cutting edge N3B node. And the difficulties all of those face were not purely constrained to the process nodes used. There were design issues and concerns with each of those mentioned product lines.

The foundries are a big drain on finances when your products aren't doing well. No one disputes that. But trying to distill this down to just *one* thing — like EUV or the lack thereof — that caused Intel to miss the boat is naive in the extreme. Everything compounds. Even if Apple only had a 1% impact, and Google had a 1% impact, and AMD, Nvidia, Qualcomm, tablets, smartphones, AI, chiplets, etc. each have a 1% impact on Intel's bottom line, compound that all together and you can easily get a 10% change or more.

Was EUV alone a 10% change for Intel, in the negative direction? Almost certainly, probably a lot more than that over time. But Intel was also trying to lead from the front, and when it lost its way at 14nm and struggled for five years, that had far reaching rammifications. The company lost the process node leadership after 14nm, where it was two years ahead, to 10nm, where it was two years behind. And that wasn't purely about EUV, it was about all the other stuff going wrong with 10nm as well.

Which is all to say, getting back to the subject at hand, that Intel made massive mistakes in the past decade. Pat Gelsinger wasn't perfect either, but if Intel was ever going to regain the lead, it needed to make massive changes. I'm not sure it's even going to keep trying to regain the lead without Gelsinger, but we'll see. Selling off the foundries as a separate entity could work out in the long run for Intel, but I'm not so sure that will be good for the US, given the current geopolitical situation. But hey, we've got that amazingly successful TSMC Arizona plant that will be great, right? LOL

Apple has zero to do with this, zero. Intel doesn't make SOC for mobile devices and Apple PC sales are only 9% of the market, Lenovo, HP and Dell dwarf them. Asus is at 7.1% and Acer is 6.6%. In the server market Apple is entirely inconsequential, virtualized x86 dominates everything with some RISC for appliances and specialized tasks.

PC vendor market share worldwide 2023 | Statista

The overall market leader in the global personal computer market in 2023 was Lenovo with a market share of 24.7 percent with HP Inc.

www.statista.com

Intel's woes are entirely centered on it's foundry business not being able to compete with TSMC and burning a ridiculous amount of cash trying. Just to highlight how bad it is, Intel is spending billions on their own Fab while using TSMC to make their flag ship products.

That's the equivalent of AnandTech paying Toms to publish articles here, while still hosting their own site.

Me talking EUV was merely highlighted the reasons why Intel was unable to compete with TSMC in the foundry business and how that impacted their slumping CPU sales. Without those sales they can't afford to keep investing into the foundry side. This is a common problem with a business model being too vertically integrated. Them missing that boat is what put them so far behind and required such massive amounts of investment to catch back up, as I highlighted earlier by them purchasing ASML's latest tech and now having to build a process node around it.

If Intel wants to continue then it needs to separate out the foundry side from the product side and allow each to develop separately. The mistakes made in the foundry side is what hurt the product side and the lack of revenue from the product side is what's hurting the foundry side. Let the foundry split off and continue developing production techniques while also getting revenue from using older existing equipment to make commodity chips that aren't in China or Taiwan, this is what GloFo does and it works. The product side can then build products using whatever vendor provides the most competitive process, TSMC, Samsung, Intel Foundry or even GloFo.

 

[thestryker]

Ehh people absolutely are and that includes TSMC.

No they really aren't. You stated "3nm" and nobody is using it for that. Did you actually read the links you posted? At this point I think you're just arguing to argue, but I cannot let blatant lies go unanswered.

If Intel wants to continue then it needs to separate out the foundry side from the product side and allow each to develop separately. The mistakes made in the foundry side is what hurt the product side and the lack of revenue from the product side is what's hurting the foundry side. Let the foundry split off and continue developing production techniques while also getting revenue from using older existing equipment to make commodity chips that aren't in China or Taiwan, this is what GloFo does and it works.

There it is you've exposed what I assumed all along: you know next to nothing about the fab industry.

The fact that no Intel node prior to Intel 3 is viable for external foundry services means this isn't a possibility, but by all means keep beating your ignorant drum.

 

[JarredWaltonGPU]

Apple has zero to do with this, zero. Intel doesn't make SOC for mobile devices and Apple PC sales are only 9% of the market, Lenovo, HP and Dell dwarf them. Asus is at 7.1% and Acer is 6.6%. In the server market Apple is entirely inconsequential, virtualized x86 dominates everything with some RISC for appliances and specialized tasks.

You keep saying "zero" but what you really mean is that it was a minor loss in the grand scheme of things. I'd agree with that.

Intel didn't make SOCs for Apple, but it made millions of x86 chips for Apple. And that was the point. Apple ditched Intel and x86, after previously ditching the PowerPC architecture for x86. When Apple came to Intel, no one disputed that it helped Intel that I can recall. And if that was true, then Apple leaving x86 and Intel did have a material impact.

Could Intel have hung on without Apple? Of course. It had done so before, it could do so again. Can Intel rise to the top of the CPU space again, without its own foundries? Sure. AMD did that recently, as I don't think anyone would dispute that its current CPUs are beating Intel in the data center and in consumer (but less so in consumer I'd say — because there's a lot more complexity than those blanket statements cover).

It's chaos theory out there, and while I'm not going to say a butterfly flapping its wings in Australia caused Intel's woes, I will say that business losses and decisions beyond the foundries and EUV have all played some part. There's almost never a single cause to most big problems, especially with a corporation as big as Intel. If Intel had gone all-in on EUV before TSMC, that wouldn't necessarily guarantee it remained at at the top. That was one big decision where it bet one way and lost, hard. But had it bet the other way, it still could have lost, because getting EUV up and running "first" was not a trivial task.

How long has Intel been working on that even now? Why isn't Intel 3, 20A (RIP), and 18A already done and beating TSMC? Because it still takes a lot of hard work, and good decisions, and on some level it does feel like Intel's culture has become a problem that it can't seem to overcome. It's why I also don't think a big cash infusion of say $30 billion would fix Intel's problems, because they're systemic.

And I would say, if anything, that's precisely why Apple left x86. Apple saw things that it didn't like, and said, "We can do better on our own. We don't need Intel." History has largely proven Apple's move to have been the correct one. And at some point, the same thing will happen to Apple, AMD, Nvidia, Google, etc. It's inevitable, in my opinion, because consistently making the right decisions is hard, and even if you do make all the best choices it doesn't always play out perfectly due to other factors that you can't fully control.

 

[palladin9479]

You keep saying "zero" but what you really mean is that it was a minor loss in the grand scheme of things. I'd agree with that.

Intel didn't make SOCs for Apple, but it made millions of x86 chips for Apple. And that was the point. Apple ditched Intel and x86, after previously ditching the PowerPC architecture for x86. When Apple came to Intel, no one disputed that it helped Intel that I can recall. And if that was true, then Apple leaving x86 and Intel did have a material impact.

Could Intel have hung on without Apple? Of course. It had done so before, it could do so again. Can Intel rise to the top of the CPU space again, without its own foundries? Sure. AMD did that recently, as I don't think anyone would dispute that its current CPUs are beating Intel in the data center and in consumer (but less so in consumer I'd say — because there's a lot more complexity than those blanket statements cover).

It's chaos theory out there, and while I'm not going to say a butterfly flapping its wings in Australia caused Intel's woes, I will say that business losses and decisions beyond the foundries and EUV have all played some part. There's almost never a single cause to most big problems, especially with a corporation as big as Intel. If Intel had gone all-in on EUV before TSMC, that wouldn't necessarily guarantee it remained at at the top. That was one big decision where it bet one way and lost, hard. But had it bet the other way, it still could have lost, because getting EUV up and running "first" was not a trivial task.

How long has Intel been working on that even now? Why isn't Intel 3, 20A (RIP), and 18A already done and beating TSMC? Because it still takes a lot of hard work, and good decisions, and on some level it does feel like Intel's culture has become a problem that it can't seem to overcome. It's why I also don't think a big cash infusion of say $30 billion would fix Intel's problems, because they're systemic.

And I would say, if anything, that's precisely why Apple left x86. Apple saw things that it didn't like, and said, "We can do better on our own. We don't need Intel." History has largely proven Apple's move to have been the correct one. And at some point, the same thing will happen to Apple, AMD, Nvidia, Google, etc. It's inevitable, in my opinion, because consistently making the right decisions is hard, and even if you do make all the best choices it doesn't always play out perfectly due to other factors that you can't fully control.

Apple is a minor player in laptop / desktop computing and not a player in server computing, they are a phone and tablet seller. I even provided the numbers showing that Asus alone produces nearly as many PC's as Apple. Pulling at strings and trying to twist pretzels is like saying Asus is responsible for Intel success / failure. Let the absurdity of that statement linger for a second.

As for Apple leaving x86, that one is easy, Hackintosh's.

Like all luxury designer brands Apple's #1 asset is it's brand image. Not it's technology, not it's people, not it's products, it's brand. Steve Jobs successfully rebranded Apple as a premium luxury designer label, and like all such labels counterfeits are a serious threat. People figuring out how to put Apple exclusive MacOS onto a non-Apple system and then release them cheaply out of China would have devalued their brand image costing them over a trillion USD. Since they already were using ARM SOC's for their main products, disposable phones and tablets, and all their software is a walled garden, switching over the smaller business unit to protect the brand image just makes sense. Having their hardware team produce their own SOC's makes it near impossible to make counterfeit Apple products running real Apple software. That alone justifies the cost.

It was never about performance, it's the hardware version of DRM.

Just watching people try to imagine the entire consumer IT world somehow revolves around a minor player is just ... wow. Apple products are luxury goods that target a very specific demographic.

Now back to Intel and it's financial woes. This has been a problem for awhile, the reason this is happening now is that Intel's net income / operating income in the last four years have been terrible and they were forced to disclose more financial info causing a panic and stock sell off.

https://www.macrotrends.net/stocks/charts/INTC/intel/net-income

https://www.macrotrends.net/stocks/charts/INTC/intel/long-term-debt

The last four years they have been borrowing heavily in an attempt to catch up. The last couple of product releases have been really bad and not enough people were buying them causing revenue shortfalls. The 14th and 2xx series haven't sold well with people mostly sticking with existing 12th or 13th platform or just going to AMD. Intel is still moving product, just not enough to keep the ship afloat with their high debt load from the R&D work associated with the fabrication plant.

This is the exact same situation AMD was in back during Bulldozer / Piledriver and they had to take drastic measures to keep the company going. There is a reason they spun off Global Foundries then eventually switched to TSMC as a parts provider, and the result of that has been the Ryzen you rave about.

 

[Kamen Rider Blade]

Old white men targeted layoff is completely false.

Brian Krzanich literally spent $300 million on Diversity Hirings.
Guess who got the boots in place?

Old White Senior Intel Engineers that went to Apple.

JarredWalton was right on the $ IMO.​

BK literally wasted INTC company funds for political DEI clout.

Look what path it sent Intel down, it sure as hell didn't help it.

ASML's low-NA EUV system used for last generation microprocessors (7nm) costs about $180 million each. Their recent high-NA EUV is required for 3nm microprocessors and costs $380 million each. The next generation Hyper-NA EUV, which should be available around 2030 and is required for sub 1nm nodes, is going to cost around $760~800 million each. Making equipment for high performance processors is ridiculously expensive, this is why all the other foundries have lagged and we just have TSMC and Samsung at the top end and even then Samsung lags behind and is very careful about their markets.

Where do you think TSMC got the $$$ to buy all the expensive ASML Low-NA EUV machines?

It was off the massive profits from changing it's ways by having Apple become it's #1 partner and investing on the best technology.

You can't be the best in the Process Node game w/o spending a massive amount of Capital Expenditure on the best machines.

You can't fund that Capital Expenditure if you didn't gain "Massive Profits" from changing the way you do Process Node Development & financial recoupment.

That's why I told you to read.

https://semiconductor.substack.com/p/the-apple-tsmc-partnership

Apple's influence on TSMC is huge.

ASML wouldn't be able to sell all those machines if TSMC didn't have the finances to buy it.

 

[palladin9479]

Where do you think TSMC got the $$$ to buy all the expensive ASML Low-NA EUV machines?

It was off the massive profits from changing it's ways by having Apple become it's #1 partner and investing on the best technology.

You can't be the best in the Process Node game w/o spending a massive amount of Capital Expenditure on the best machines.

You can't fund that Capital Expenditure if you didn't gain "Massive Profits" from changing the way you do Process Node Development & financial recoupment.

That's why I told you to read.

https://semiconductor.substack.com/p/the-apple-tsmc-partnership

Apple's influence on TSMC is huge.

ASML wouldn't be able to sell all those machines if TSMC didn't have the finances to buy it.

I think you really do not grasp the size of TSMC, they could of easily afforded those machines, easily. TSMC and ASML have been collaborating in this way since 1970's, go check that if you don't agree with me.

If Apple didn't give that money to get first pick of the TSMC's next generation process nodes, someone would have or they would of just did it themselves. All Apple did was buy the rights to have access to all the newest process nodes, they pay a hefty premium to have first pick and never worry about supply shortage. Since they are a premium luxury brand maker, it makes sense for them to spend heavy amounts of money to guarantee supply and just pass that cost to the consumer, aka $700 wheels.

People who like to own luxury brands are not only willing to pay big money for those brands but assume the bigger value makes them more special.

 

[palladin9479]

Guys ... do people not realize how big the companies involved here are...

Apple is a 3.7 trillion USD company. TSMC is a 1 trillion USD company (careful of conversion from TWD). In comparison Intel is a mere 96 billion USD company, "Money Money nVidia" is 3.4 trillion USD and ASML is 266 billion USD.

The weird one is Samsung Electronics at 243 billion USD. We think of it as a single company but it's not, it's part of the Samsung Group which is a massive Korean conglomerate that makes everyone else except TSMC look small. These guys do defense, construction, insurance, finance, medical, biotech, pretty much everything that makes money and is controlled entirely by a single family.

 

[UnforcedERROR]

TSMC and ASML have been collaborating in this way since 1970's, go check that if you don't agree with me.

I know very little about this space but a quick glance tells me both companies were founded in the 80s (though ASM was started in the late 60s it seems). I'm welcome to being wrong here.

 

[palladin9479]

I know very little about this space but a quick glance tells me both companies were founded in the 80s (though ASM was started in the late 60s it seems). I'm welcome to being wrong here.

No your right though it gets a bit technical. TSMC proper was founded in 87, prior it was a part of ITRI.

https://en.wikipedia.org/wiki/Industrial_Technology_Research_Institute

The collaboration started in the 70's when Immersion lithography was first proposed, engineers in ITRI worked with engineers in ASML to see if it could be done. Once it was possible they spun off a dedicated company, TSMC, to do it. TSMC was initially funded by the Taiwan government, which is still the largest stake holder at 6.4%.

Who Owns TSMC: The Largest Shareholders Overview

Visual overview of who owns Taiwan Semiconductor Manufacturing Company Limited (TSM) and who controls it. With a list of the largest shareholders and how much is each of their stake worth.

www.kamilfranek.com

And while TSMC is a public company, the Taiwan government views it as a national security asset. This is why it's utter nonsense to think TSMC somehow "needs" money. They have the backing of a national government.

 

[zypha]

Lisa Su had basically nothing to do with AMD's turn around. While she was handed a company in worse financial shape than Intel, all the major components that turned the company around were already complete are well under way by the time she took over. All she had to do was not screw it up. Compared to Intel, AMD looks good. Compared to Nvidia, their other major competitor, Su looks like a complete failure. Gelsinger is forced out partially because Intel missed the AI bus? AMD was in a much better position to jump on the AI bus and completely missed it too.

That's unfair to Su.

AMD didn't just turn itself around; it also grew in several key markets. Hundreds of decisions had to be made along the way, and she deserves credit for making them.

 

[Kamen Rider Blade]

And while TSMC is a public company, the Taiwan government views it as a national security asset. This is why it's utter nonsense to think TSMC somehow "needs" money. They have the backing of a national government.

Just because TSMC is a "National Asset", doesn't mean it has a endless pocket book.

No Government likes dumping $$$ to prop up companies if they don't have to.

Just look at all the finangling that the US Government is doing over dolling out money for the CHIPs act to Intel.

Apple is a 3.7 trillion USD company. TSMC is a 1 trillion USD company (careful of conversion from TWD). In comparison Intel is a mere 96 billion USD company, "Money Money nVidia" is 3.4 trillion USD and ASML is 266 billion USD.

Those are just stock valuations currently.

They fluctuate incredibly with the price of the stock.

The actual value of the company is far harder to measure.

And most of that valuation, isn't Liquid Cash.

 

[Kamen Rider Blade]

I think you really do not grasp the size of TSMC, they could of easily afforded those machines, easily. TSMC and ASML have been collaborating in this way since 1970's, go check that if you don't agree with me.

If Apple didn't give that money to get first pick of the TSMC's next generation process nodes, someone would have or they would of just did it themselves. All Apple did was buy the rights to have access to all the newest process nodes, they pay a hefty premium to have first pick and never worry about supply shortage. Since they are a premium luxury brand maker, it makes sense for them to spend heavy amounts of money to guarantee supply and just pass that cost to the consumer, aka $700 wheels.

People who like to own luxury brands are not only willing to pay big money for those brands but assume the bigger value makes them more special.

Nobody else had the $ to invest into TSMC and jump start profits to buy the EUV machines.

If you don't believe it, then read this.

https://quartr.com/insights/company...ire-tsmcs-revolution-and-morris-changs-legacy

ASML’s EUV lithography machines – measuring the size of an American school bus and weighing 150 tons – are sometimes referred to as the most complex machine on earth, and without TSMC, Samsung, and Intel, the technology might not have become commercially available. Or, available might be a strong word, it takes immense R&D and CapEx investments, operating efficiency, and scale to be able to operate an EUV machine profitably. Not to mention the price tag of over $200 million a piece.

This means that only a tiny number of companies globally can rationalize investing in these machines. Three of those are TSMC, Samsung, and Intel, which all three helped fund ASML’s last big R&D push to develop EUV in 2012, in exchange for an equity stake without voting rights each. Despite bringing the largest amount to the table back in 2012, Intel could not keep up with the pace over in Asia and did not start using EUV until 2023. TSMC and Samsung were years ahead, already producing chips at scale using ASML’s breakthrough technology.

Intel couldn't afford it, only TSMC had the financial CapEx to buy all those EUV machines along with a customer willing to help bankroll it and work through any issues on leading edge nodes.

It's not like there are that many companies who can afford to dump $$$ into being first on the latest nodes. Apple is that company.

Despite what you may think or feel, if you blindly believe that TSMC could do it by itself, you are sorely mistaken.

There has to be a steady customer with a pocket book that is willing to help bankroll this type of high cost CapEx.

W/O that, the Process Node development would go FAR slower and the amount of machines bought would be slower, which is a compound effect on how many machines can be bought in bulk which affects the costs of the machines themselves due to bulk order pricing on a global scale.

 

[NinoPino]

No your right though it gets a bit technical. TSMC proper was founded in 87, prior it was a part of ITRI.

https://en.wikipedia.org/wiki/Industrial_Technology_Research_Institute

The collaboration started in the 70's when Immersion lithography was first proposed, engineers in ITRI worked with engineers in ASML to see if it could be done. Once it was possible they spun off a dedicated company, TSMC, to do it. TSMC was initially funded by the Taiwan government, which is still the largest stake holder at 6.4%.

Understand your point of view. Anyway, saying that ASML and TSMC collaborate from the '70 is wrong in any sense. ASML and TSMC was founded a decade after and is simply a nonsense to say that a company collaborate with another company before its foundation.

Also saying TSMC spun off from ITRI is something like saying Intel spun off from Berkley.

 

[ottonis]

Just a speculation, but it seems that Mr. Gelsinger, who is definitely a very smart guy and incredibly versed engineering expert, fell over current massive losses (16 B USD just in the last quarter) and that the suits in the boards didn't want to lose their bonuses, as they were waiting for Gelsinger's strategy to turn Intel into profit zone again.

My guess is that Intel will most certainly abandon the manufacturing branch just like AMD did with GlobalFoundries, and that this alone would set save Intel billions of dollars of salaries. This will probably put them back into the profit numbers and the suits will get the bonuses and everyone will be happy.
That being said, Intel still does not seem to have a clear and convincing strategy for AI/GPU....

 

[passivecool]

Pat was no spring chicken.
He'd been around the block 50 times.
He came with open eyes
and a mouth full
of promises
and self praise.
And a head full of strategies,
outdated.
And
He
Failed.
So he can take his quarter billion,
or whatever his severance will be
stuff it into his pillow and
cry into that.

Maybe he should have prayed for wisdom, instead, and not taken the job.

 

[Kamen Rider Blade]

My guess is that Intel will most certainly abandon the manufacturing branch just like AMD did with GlobalFoundries, and that this alone would set save Intel billions of dollars of salaries. This will probably put them back into the profit numbers and the suits will get the bonuses and everyone will be happy.
That being said, Intel still does not seem to have a clear and convincing strategy for AI/GPU....

They signed the chips act, so they can't.

Even if they IPO it, they have to own at least 50.1% of the shares in perpetuity.

So they HAVE to make the foundry work.

Unlike AMD, which had the option to go their seperate ways.

 

[JarredWaltonGPU]

I think you really do not grasp the size of TSMC, they could of easily afforded those machines, easily. TSMC and ASML have been collaborating in this way since 1970's, go check that if you don't agree with me.

If Apple didn't give that money to get first pick of the TSMC's next generation process nodes, someone would have or they would of just did it themselves. All Apple did was buy the rights to have access to all the newest process nodes, they pay a hefty premium to have first pick and never worry about supply shortage. Since they are a premium luxury brand maker, it makes sense for them to spend heavy amounts of money to guarantee supply and just pass that cost to the consumer, aka $700 wheels.

People who like to own luxury brands are not only willing to pay big money for those brands but assume the bigger value makes them more special.

You're REALLY stretching here. TSMC is worth about $1 trillion today. But we're talking about the state of things when Apple really started putting money into the pot. The first Apple-designed SOCs were the Swift architecture in the A6 in 2012, but the first cutting-edge node used by Apple was the A12 on N7, released in 2018. Which means it was being designed throughout 2017, probably even 2016 and 2015.

So, in 2017 when TSMC's market cap was only $150 billion — still a lot, yes, but actually slightly less than Intel at the time — Apple invested heavily into being first on N7. Apple incidentally was 'only' worth about $750 billion at the time. And that's a key factor: Apple was worth about five times as much as TSMC. (And Nvidia was worth around $100 billion, while AMD was worth a paltry $10 billion — the wonderful pre-Ryzen era!)

How much did that investment in the first N7 wafers cost Apple, and how much did it benefit TSMC? Apple was probably telling TSMC, "You get EUV working and we will help bankroll it" at the time. Definitely not "nothing." Remember, this is the same N7 node that Nvidia rejected for its own use two years later because of its high cost! It went with a clearly inferior Samsung 8N node for Ampere. So in 2017, the investment in wafers and design by Apple was far from trivial. Or do you think that AMD, worth just $10 billion, was putting in the money for N7 ramp-up that TSMC needed? 🤔

And the idea that Apple, as a regular user of TSMC's cutting-edge nodes for the past seven years, isn't actually helping at all to work out the kinks is ludicrous. That's the same as saying that Nvidia hasn't helped refine things at all with the various nodes it has used. When a corporation worth billions of dollars invests heavily in a particular node and even comes up with a tuned and tweaked version of said node, it is absolutely helping. Or I suppose by your logic Intel's design team for CPUs hasn't helped at all with the foundry side of things for the past several decades — even though it was the primary driver of the foundries.

Ergo: If TSMC made moves that hurt Intel, by extension Apple, Nvidia, AMD, and even Intel itself has helped TSMC and thereby hurt Intel.


PS: Your calling Apple a luxury brand as some sort of insult is a red herring. It's completely irrelevant whether Apple is "luxury" or not. What's relevant is the amount of money it spends on R&D and product design. "Oh, it's a luxury brand! That means it has no idea how tech works, even though it's the (currently) most valuable company in the world!" 🙄 Give it a rest.

Apple isn't perfect, but it sure as heck knows a lot about technology and has contributed to the greater internet/technology/etc. space in a lot of ways — some of them admittedly bad, because it is Apple and it loves to do proprietary and unnecessary solutions as much as anyone. But it's not like Apple is an oil conglomerate (eg, Saudi Aramco) or retail store chain (eg, Walmart). It operates directly in the technology space, just like Nvidia, Intel, AMD, Qualcomm, Samsung, and others.

Labeling it "luxury" and then dismissing it because of that label is a super weak stance. Hate Apple because of vendor lock-in or arrogance all you want, but it's still a major tech company. And it's worth over 3X as much as TSMC right now. You can keep saying it has done "nothing, ever, forever and ever nanananabooboo" to impact Intel. That's a nonsense stance. Intel may have largely dug its own holes, but the success of other companies, who were working directly and in a large way with Apple, means there is a relation there that can't simply be ignored.

 

[COLGeek]

Just a nudge to the community, this thread has meandered quite a bit from the original topic by huge leaps and bounds. It might be a good idea to focus on the actual topic, vice an on-going dialogue of the entire tech industry.

Thank you.

 

[phead128]

The Apple-EUV angle is a bit overplayed since TSMC was making 7nm on DUV only with decent yields (Before later switching to EUV only for 7nm), there is no excuse why Intel couldn't do the same on 10nm/7nm on DUV only except that it was simply bad at execution, with or without EUV funded by Apple.

Also, there is the example of Samsung achieving leading edge nodes without Apple as a customer. SMIC can even achieve 7nm on DUV without Apple as a customer. So Apple helps, but there are plenty of examples of other fabs achieving leading edge without deep pockets of Apple. More money won't fix Intel's problems which are rooted in poor leadership and poor execution.

 

[ex_bubblehead]

As this has gone FAR into left field and is no longer addressing the topic in question the time has come to close this thread and put it to bed.