• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Discussion Intel's past, present and future

W

whoshere

Member
Looks like he's been unable to right the ship and has been ultimately ousted.

What do you think the future of Intel will look like?

Source:

www.intc.com

Intel Announces Retirement of CEO Pat Gelsinger

www.intc.com www.intc.com

This thread has been moved to P&N due to some posters responding to its political components. P&Ners, please don't comment on the move, just the substance of the situation. CPUs and Overclocking members, welcome to the wider world we all live in these days.
Perknose
Forum Director
 
Last edited by a moderator:
Those ****** posted this the day before Battlemage…

First off Intel needs a new board of directors and a CEO with full control of the company breaking thru its existing political barriers. It’s going to hard. The foundry depends on 18A.
 
poke01 said:
First off Intel needs a new board of directors and a CEO with full control of the company breaking thru its existing political barriers. It’s going to hard. The foundry depends on 18A.
Click to expand...
It's a publicly traded company.
They're not doing caudillo stuff lol.
 
This is a surprise but not a complete surprise considering the history of the BoD's decisions over the past few decades.

We're likely witnessing the history of the fall of this company. Actually it has been set in stone years ago. Otellini started the whole failure of negotiations with Apple, because he had no vision, when it would have been obvious to any one of us. Maybe he did have vision, but wasn't enough to defeat his obsession with numbers, aka MBA. Specifically said that he didn't think the "volume would be there".

Then Kraznich was incompetence compiled with a healthy dose of psychopathy.
 
DavidC1 said:
This is a surprise but not a complete surprise considering the history of the BoD's decisions over the past few decades.

We're likely witnessing the history of the fall of this company. Actually it has been set in stone years ago. Otellini started the whole failure of negotiations with Apple, because he had no vision, when it would have been obvious to any one of us. Maybe he did have vision, but wasn't enough to defeat his obsession with numbers, aka MBA. Specifically said that he didn't think the "volume would be there".

Then Kraznich was incompetence compiled with a healthy dose of psychopathy.
Click to expand...

It was not "obvious" to any of us. Anyone here who claims they knew back in 2005 when Apple was only rumored to be working on a phone that it would be a massive success is a liar. It is completely believable that he didn't believe the volume would be there from Apple making phones. When Steve Jobs said at launch he could see Apple eventually taking 1% of the cell phone market (meaning nearly 10 million phones a year) most people thought his reality distortion field was affecting him.

Intel wanted Apple to use x86, Steve Jobs said "absolutely not". That's why the negotiations with Apple failed. Intel wasn't a foundry and didn't have any aspirations of becoming one, not 20 years ago. So the only way a deal with Apple could have happened would be with StrongARM. The "x86 everywhere" sentiment within Intel was probably too strong for that to be an option. Plus they knew they would effectively become a foundry because Apple didn't want a CPU, they wanted a full SoC. Not sure if any of the StrongARM/Xscale lineup would have qualified.

We can fault them for not being serious about becoming a foundry in their first attempt, and they should have moved heaven and earth to secure Apple as a customer then when they were well known to be unhappy working with Samsung. If they'd teamed up with Intel instead of TSMC, the last decade would have been very different. Not sure we should blame the board, CEO, or VPs/upper middle management for that failure though.
 
I think we'll see Pat's gambles pay off over the next 5 years. Hopefully 18A turns out as strong as he wanted it to, and we start seeing some seriously competitive products from Intel on it. I don't think we'll see big external customer buy in until Intel has won their trust again, but maybe we'll start to see that post-18A if it works out.

Given they've appointed co-CEOs, I think the company is going to be split into two parts. I think Intel Product is going to be spun off (because x86 is tied to a stagnant PC market), and Foundry stays in house as the potential bigger money maker long term.
 
NTMBK said:
I think we'll see Pat's gambles pay off over the next 5 years.
Click to expand...

Even if there was a 50% probability that Pat's gamble will pay off, Pat would still be the CEO.

The problem with the strategy to become a full foundry is that it needs a steady inflow of cash to fund the foundry operations, and Intel just does not have it.

The smoke and mirrors Pat tried, by selling equity stakes etc is only a bandaid.

To fully fund the foundry - takes constant spending of money (on R+D and equipment). So the BoD was facing a choice in going full steam and running out of money (= BK) or changing course.

But unfortunately, there are no easy and obvious answers on what to do next. Only one thing is clear: Intel can't afford to fund the foundry alone.
 
Never mind the mad "x86 everywhere" mentality. Mobile and smartphones were all about volume, volume, and volume.

Intel - with a CPU design which was frankly an embarrassment in 1981 with the 8086 with a horrendous memory module, registers etc. - got where they were not because of any technical merit but rather due to volume, volume, and volume.

The high margins, low volume RISC players all lost to Intel.

Problem with the Intel distortion field was, that Intel's hybris was such that they actually believed that it was x86's technical merit rather volume which made them big.

Hence them ignoring mobile. The huge volume of mobile allowed TSMC to get ahead and stay ahead.
 

According to Bloomberg, Pat was given the option to retire or be removed, and chose to announce the end of his career at Intel. Enough said...

The departure of Gelsinger could lead to more dramatic strategic shifts.

Yep, after 5NI4Y plan, Intel might scale down IFS and goes back to TSMC.... I really hope people won't believe 18A will save Intel cause if upcoming NVL's compute tile is going to be fabbed by TSMC's N2 process, IFS has no way catching up to TSMC's standard, just like SF which has been trying with their SF3E, SF3, SF2 which sound more advanced than TSMC's process...
 
Not gonna lie I had to Google what 18A referred to. I'm so out of the loop on chip process stuff, figured it was something related to scaling and HPC.

Gelsinger’s departure comes as Intel’s financial woes have been piling up. The company posted a $16.6 billion loss and halted its dividend in the most recent quarter, and its shares have fallen by about 60% since he took over as CEO.
Click to expand...

Ouch, I was actually under the impression they were doing a bit better since 2021. I suppose with machine learning being at the forefront of modern computing, there's no competing with NVIDIA.
 
Whats almost deliciously schadenfraudistic is Pat was a solid engineer back in the day. When he came back to Intel, he seemed like he was going to make some real changes...
One would think an Engineer would have had a real close look at what the competition was actually doing and not risk looking like an idiot for something that was a massive money saver and key to system/process wins like chiplets.
The rearview mirror comment was a bit sad as it was so obviously farcical even at the time. They really need/ed to bring in someone from outside Intel who hasn't coasted on past succes but comes from an industry/company that has had to fight to survive. Really might be too late for Intel to avoid the Xerox/Kodak end game.
 
Last edited:
Intel built an empire on maybe the largest reality distortion field of all time. It's easy to be seen as a technological powerhouse when you have little to no competition and Intel instinctively knew this that's why they resorted to bribing partners to keep AMD from reaching critical mass. Eventually that plan failed AMD become more than competitive. Apple went their own way, TSMC became state of the art in fabrication, Nvidia became the largest market cap company on earth doing what Intel could never dream of.

The only thing keeping Intel going is decades of "never got fired for buying Intel" and that is on life support.
 
Intel is studied as an example of a double-moat company, and how those reinforced each other. x86 was (mostly) proprietary to the company, and it was exclusive to Intel's process. So if you wanted the fastest chip you bought x86, and if you were captured by x86 (Microsoft) you funded their process. Each moat gave value to the other, and together they were nearly impossible to beat. The downside is that you can't invite other architectures onto the process that compete with x86, because that undermines the moat structure, which limits your ability to scale.

Intel had 3 problems to manage:
1) DOJ breaking that exclusivity to x86 to some degree
2) They could never stumble on process (and they did)
3) The market couldn't scale even within x86 because making x86 suitable for the PC/server market forced it to be unsuitable to the low power market.

Two developments outside of Intel counter these moats:
1) ARM license their IP to pretty much any comer, so anyone can break out of the Intel SKU structure to develop silicon for narrow use cases. This is why Apple cofounded ARM to begin with, for such a narrow use case, and they returned to the architecture (with founders license intact) for their mobile products. This allows Apple (and anyone else) to make silicon that meet their exact requirements rather than picking from Intels menu of offerings.
2) TSMC learns the originating lesson from Intel that Intel lost sight of. Semi manufacturing has very low marginal costs and very high capital costs. As such, scale is everything, and Intel cut off their scaling opportunities. But TSMC could collect customers to cover those increasing costs to tip up new nodes. By the time Intel were showing real problems, Apple was spending more on silicon than Intel was, let alone all of the other TSMC customers - AMD, Nvidia, Qualcomm, Broadcom, etc. Players like Apple have options as they can always jump to Samsung (as they jumped from Samsung) if TSMC stumbles. Intel had no such option.

Intels double moat ultimately trapped them. They couldn't turn into a foundry without breaking that. And as soon as they stumbled on process, that was pretty much it. Apple showed the benefit of a licensable IP that they could tailor to their needs along with a competitive foundry space where they could always be on the best process. AMD dumped off GlobalFoundries to gain the same benefits.

The only path I see is for the US government to recognize that if they want a competitive domestic foundry for ITAR work (particularly as DOD is leaning into AI more), they're going to have to build a ULA-like entity around Intels fabs, with buy-in from the US customers - Intel, Apple, AMD, etc. That would stabilize thing and ensure domestic owned capacity. Even if this entity wasn't quite competitive with TSMC, there are plenty of products they make that would be suitable - Apple's S series, their new radios, etc. Between these entities they would be able to provide enough volume for this entity to be reasonably competitive.
 
fkoehler said:
Whats almost deliciously schadenfraudistic is Pat was a solid engineer back in the day.
Click to expand...

Being a good engineer does not make one a good CEO of a multi billion dollar company. Being "just" an engineer without any real aptitude for management or finance makes one no more qualified to be CEO than does holding a degree in management or accounting.

Tech people love to bag on MBAs as if they are responsible for all the ills of companies, but just because you understand the business a tech company is involved in in depth does not make you any more qualified to manage it than someone who couldn't tell a FinFET from a fiber transceiver. That's why a lot of small startups never get past that "small" stage, because the founder is often someone who is really smart in their field and thinks that alone makes them qualified to grow the business past the point where the guy on the top no longer has time to delve into technical matters but is solely a manager.

Yes there is no shortage of examples of engineer lead companies that took a big turn for the worst when the MBAs/beancounters took over, and these days Boeing is probably a prime example. But the problem with Boeing wasn't that they didn't have an engineer in charge. It was that when McD came in an basically took over management they pushed out ALL the engineers for non-engineers all over the structure, and didn't listen to the engineers who told them what they didn't want to hear when they had all their great "financial engineering" ideas like spinning off their airframe manufacturing.

There was a lot of consternation when Tim Cook took over for Steve Jobs as CEO of Apple. Steve Jobs was a visionary, and that's a role that's impossible to replace, but he'd learned a few things about being a successful manager since he was fired from Apple the first time around. We may not agree with all of his methods, but it is hard to argue against his results. So many people believed that Tim Cook, a nerdy MBA with an "industrial engineering" degree (which isn't what people here would think of as an "engineer", they are the beancounters who drive efficiency in production which is why he was COO under Jobs) would guarantee that Apple's best days are behind it. The stock has gone up by 10x since then. You can quibble that Apple isn't "innovative" any longer but compared to who? What has Google innovated in the last decade, other than finding more ways to collect your personal information and sell ads? Ditto for Facebook and Microsoft. Intel sure hasn't. Nvidia has, with the rare example of a founder who was not only technically brilliant but also turned out to be a terrific CEO.
 
One good thing about ARL and later client products is that they can be flexible with the nodes.
 
Doug S said:
It was not "obvious" to any of us. Anyone here who claims they knew back in 2005 when Apple was only rumored to be working on a phone that it would be a massive success is a liar. It is completely believable that he didn't believe the volume would be there from Apple making phones. When Steve Jobs said at launch he could see Apple eventually taking 1% of the cell phone market (meaning nearly 10 million phones a year) most people thought his reality distortion field was affecting him.

Intel wanted Apple to use x86, Steve Jobs said "absolutely not". That's why the negotiations with Apple failed. Intel wasn't a foundry and didn't have any aspirations of becoming one, not 20 years ago. So the only way a deal with Apple could have happened would be with StrongARM. The "x86 everywhere" sentiment within Intel was probably too strong for that to be an option. Plus they knew they would effectively become a foundry because Apple didn't want a CPU, they wanted a full SoC. Not sure if any of the StrongARM/Xscale lineup would have qualified.

We can fault them for not being serious about becoming a foundry in their first attempt, and they should have moved heaven and earth to secure Apple as a customer then when they were well known to be unhappy working with Samsung. If they'd teamed up with Intel instead of TSMC, the last decade would have been very different. Not sure we should blame the board, CEO, or VPs/upper middle management for that failure though.
Click to expand...
To expand on @johnsonwax post on Intel being a double-moat company, I'd make the case that Apple's inquire was the final turning point to that.

The situation was pretty innocent back then: Intel had StrongARM/XScale as a sideshow and was already known as a foundry powerhouse. So to Apple, riding on its huge iPod successes, Intel made for an ideal partner.

Generally at the time the possibility for an Apple phone was linked to iPods already being a success (as a result public guesses on how a potential Apple phone would look like were very funny back then). In 2005 iPod sales were strong on the rise, and sales would peak around 2008, one year after iPhone was revealed.

So Intel faced the choice:
  • Either breaking the double-moat and serve Apple as a customer with its XScale product using its foundry. With hindsight this would have been the best possible choice for Intel, allowing its foundry to organically grow into a service open for external customers.
  • Or reinforcing the double-moat by fully focusing on x86. That was the obvious choice and what Intel did. Consequently on June 27, 2006 Intel sold the biggest chunk of XScale to Marvell.
In Apple Intel had a customer lined up for an expanded business opportunity. Intel would subsequently repeatedly try to expand its business on its own and fail again and again to this day.
 
Charlie has more thoughts:

www.semiaccurate.com

Why Did Intel Fire CEO Pat Gelsinger?

Yesterday the Intel board did the idiotic and fired CEO Pat Gelsinger.
www.semiaccurate.com www.semiaccurate.com

Given he seemed quite fond of Pat from the start (and definitely had some moles in Intel) it's worth reading.

Some snippets:

The first problem was that they knew it would not be ready on that date, not meet the intended specs, and usually wouldn’t be close. Design teams knew the other side was lying but what could they do? A few years later the process was indeed late, occasionally partially working, and met every letter of the law that governed bonus structures. Designers would then force a few working devices out to an OEM so that their bonuses, paid if device X shipped in quarter Y, did ship then. Sure yields were financially untenable but their new BMW had heated seats.


This isn’t to put the blame solely on the process side of the company, everyone lied. One great example was when Tim Cook met with Intel folk over their cellular modems. He directly asked someone I won’t name, “Will it be ready in time?”. The Intel exec said, “Yes”. He was lying, everyone on the Intel side knew he was lying but didn’t contradict the boss. From what we understand, Tim Cook also knew he was lying, and we know several Apple personnel in the room definitely knew it was well past a fib. If you have read this far, you understand how that program, and later the entire Apple/Intel relationship ended. It was for cause.
Click to expand...

...

With that in mind, if you were aware of Intel and the issues they had in the 1990s, there were period where things taped out but didn’t post, errors upon errors, and at times they built up into critical pain points. Intel tasked a young engineer to fix the process, procedure not semiconductor, and he did. He put procedures in place to make sure the costly errors that were plaguing Intel could not happen or at least were either very rare and caught early on. Any guesses as to who that engineer was? Hint: His first name was Pat.

From that point on, Intel became an execution machine and it showed in marketshare, financial results, and just about every other relevant metric. Compare and contrast this to AMD in the Barcelona days. I won’t get into personal conversations but after the third attempt at launching a working version of that part, I had a conversation with the person in charge of that chip. I related the story about how Gelsinger put procedures in place to make sure things like what AMD was seeing didn’t reoccur. This person looked me in the eye and said, “We couldn’t have forseen these problems, they were human error”. At that very moment I knew AMD was dead, and they effectively were for a decade. What did AMD do? They promoted the guy. It wasn’t until a radical management change took place almost a decade later that the culture was forced to change and, well, AMD is now beating Intel like a drum.
Click to expand...

...

To make matters worse for the company, the question that SemiAccurate has been debating with many financial folk for the past few years just came to the fore. “If not Pat, who?”. This isn’t a slight against Mr Zinsner or Ms Johnston Holthaus, it is rhetorical. The one name that the board was trying to get before they settled on Gelsinger is not going to leave her current company so that exhausts SemiAccurate’s list of qualified candidates. There may be others but, well, good luck to them. Sure Pat cleaned out the big skeletons in the closet, took the beating for the financial mess that wasn’t his doing, and cleaned up the culture, but was that enough for the next chump who steps in to the position to survive? We will see.


So that brings up the big question, what caused the board to fire Pat? Yes we know that officially he ‘retired’ but, well, he didn’t. They summarily canned him and didn’t have the guts to own up to it. Either way, why? Well after some digging, SemiAccurate was told the reason and it is, err, stupid.
Click to expand...
 
Last edited:
Bigos said:
I am not reading an ad for a full article.
Click to expand...
Well there is still relevant info there. Mainly that the board did fire him (which is the messaging already from quite a few places).

Connecting the dots it does indeed seem that they wanted more AI-AI-AI-AI-AI, and Gaudi 3 sales and GPU roadmap wasn't to their liking.

IMO the entire board should be replaced
 
fkoehler said:
Whats almost deliciously schadenfraudistic is Pat was a solid engineer back in the day. When he came back to Intel, he seemed like he was going to make some real changes...
One would think an Engineer would have had a real close look at what the competition was actually doing and not risk looking like an idiot for something that was a massive money saver and key to system/process wins like chiplets.
The rearview mirror comment was a bit sad as it was so obviously farcical even at the time. They really need/ed to bring in someone from outside Intel who hasn't coasted on past succes but comes from an industry/company that has had to fight to survive. Really might be too late for Intel to avoid the Xerox/Kodak end game.
Click to expand...
Pat may well have had the right idea with the 5NI4Y strategy; however, it was more than a day late, and way more than a dollar short.

Furthermore, Pat failed to understand that the end of Moore's Law would come from financial limitations, not technical ones.

These fundamental issues were missed by Pat (and many industry on-lookers).

I have been posting for months that the current Intel financial drain was not only not sustainable, but not palatable by the board for the length of time that the drain would necessarily be required to continue before the turn-around of the foundry and the re-tooling of the foundry to become an external supplier could start paying dividends.

I admit, I suspected that Pat would be around until late 2025 when even the successful implementation of 18A would show conclusively that even the technical success was yielding a financial failure.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top