News Three senior execs to retire from Intel Foundry, including respected semiconductor veteran Gary Patton

[Admin]

Three senior Intel Foundry executives are retiring as part of a major leadership reshuffle amid restructuring and uncertainty over the 14A process.

Three senior execs to retire from Intel Foundry, including respected semiconductor veteran Gary Patton : Read more

 

[acadia11]

Bout ….. time!based as someone astutely pointed out Lip is the Hector Ruiz of Intel he’ll probably just promote someone … instead of looking outside ir poaching someone from TSMC or Samsung … and continue the inbred farce they call a real attempt to shake up Intel other than cutting cost….

 

[atomicWAR]

As Intel bleeds more and more veteran talent... I find it increasingly unlikely they can dig themselves out of the pit they jumped head first into in the pre-zen era. I hope they turn around things soon for the sake of competition but unfortunately I find that more improbable by the day.

 

[rluker5]

Were these old executives good at their jobs? We know Intel foundries have not been doing the best for a while. It could be that these 3 were somewhat responsible for that.

Or they could have been great and will be dearly missed, or somewhere in the middle.

Without more information, judging whether this is good or bad news is premature.

 

[acadia11]

How the … could they be great Intel foundry or lack there of has been a disaster since it became cost prohibitive to develop leading edge tech and only be a supplier to yourself in the mid to early 2010s? They’ve been late with process node shrinks since before EUv and were late to EUV scene causing them to have to seek out TSMC for support in 2017 which increased the cost of their chips as well. This shake up needed to happen years ago … and I can’t see anyone from their current era being any good for the company. Think about it if you were on the board this 15 year spiral happened on your watch, gotta go. If you were an executive and had good ideas and weren’t to blame you are to blame because you don’t have the weight necessary to get isht done even if your ideas were the right ones you couldn’t push anybody to do them, which means you aren’t an effective leader. Which means got to go. The baby and the bath water needs to go … if anyone needed someone walking out with a kitchen sink … it’s Intel.

 

[bit_user]

Were these old executives good at their jobs? We know Intel foundries have not been doing the best for a while. It could be that these 3 were somewhat responsible for that.

It sounds like Garry Patton was key to the IFS strategy and came along too recently to be blamed for execution failures on their process roadmap. Don't know about the others.

It's indeed possible that some house cleaning was necessary, in at least some of these cases. Getting rid of one of the guys key to the IFS strategy does indeed seem like a cause for concern, but I guess we won't know if it's going to affect IFS' business plan unless it does.

Some of these departures might've been on the initiative of the execs, who simply don't have faith in the company's direction. They're not all necessarily being fired or restructured out of jobs.

 

[ezst036]

If your experience is taking the ship down a few notches, then "well respected" because of "your experience" is not valuable.

I can't speak personally about Gary Patton or any others, just making this observation. This or some in particular might be a bad move.

However not all experience is good experience. Not all expertise is good expertise. Things change, times change, people change.

 

[bit_user]

If you were an executive and had good ideas and weren’t to blame you are to blame because you don’t have the weight necessary to get isht done even if your ideas were the right ones you couldn’t push anybody to do them, which means you aren’t an effective leader. Which means got to go.

So, you're saying that even people down in the engine room should be held responsible for the sinking of the Titanic? This amount of victim-blaming is laughable, if it weren't so sad.

I guess you've never worked in a big company, because there are limitations on the power and influence of an exec, no matter whether or not they're an effective leader. Their primary responsibility is to keep their own house in order and execute to their objectives. If an exec tries to overstep their bounds too often or by too much, they get shown the exit by their peers and superiors. I just saw this happen at my company, at the end of last year, when my boss' boss got fired for stirring up too much trouble elsewhere in the org.

There's a saying that comes to mind: "a fish rots from the head down". The corporate culture get established and reaffirmed at the top. Now, if the corporate culture highly prioritizes managers and employees taking a defensive posture and not speaking up about problems or taking initiative, at some point that begins to select for which employees stick around and get promoted. When people who embody that mentality become the workforce, or at least the management, it does get a little harder to say they're not at fault. Therefore, because a culture problem is at least partly a people problem, changing such a culture will necessarily involve breaking some eggs.

So, I'm not trying to let the execs, managers, or even rank-and-file employees off the hook. I'm just saying they're probably not the root of the problem and we just don't have enough information to know which of them really need to go. The one thing I'm certain about is that a lot of people are getting let go who were not bad apples, and it's just lying to make ourselves feel better to say otherwise.

 

[DS426]

The departures come as Intel implements a major cost-cutting plan, aiming to reduce its global workforce by 15%. The company expects to close the year with approximately 75,000 employees worldwide, which means that the company will have fired 30,000 people in 2025.

Isn't the goal a 22% workforce reduction now?

https://www.reuters.com/business/intel-is-cutting-more-jobs-ceo-tan-tries-fix-manufacturing-missteps-2025-07-24/

Using "fired" very loosely there; some of it is attrition and not rehiring, some encouraging retirement and maybe some forced early retirements, voluntary employer changes (might as well get ahead of the axing, right?), etc. The amount of actual layoffs is probably somewhere around 75% of the total workforce reduction numbers, but I suppose we won't really know for sure until the dust settles in 2026.

 

[pug_s]

As Intel bleeds more and more veteran talent... I find it increasingly unlikely they can dig themselves out of the pit they jumped head first into in the pre-zen era. I hope they turn around things soon for the sake of competition but unfortunately I find that more improbable by the day.

I might disagree with you. They need another Lisa Su and some new blood because Intel just can't innovate.

 

[acadia11]

So, you're saying that even people down in the engine room should be held responsible for the sinking of the Titanic? This amount of victim-blaming is laughable, if it weren't so sad.

I guess you've never worked in a big company, because there are limitations on the power and influence of an exec, no matter whether or not they're an effective leader. Their primary responsibility is to keep their own house in order and execute to their objectives. If an exec tries to overstep their bounds too often or by too much, they get shown the exit by their peers and superiors. I just saw this happen at my company, at the end of last year, when my boss' boss got fired for stirring up too much trouble elsewhere in the org.

There's a saying that comes to mind: "a fish rots from the head down". The corporate culture get established and reaffirmed at the top. Now, if the corporate culture highly prioritizes managers and employees taking a defensive posture and not speaking up about problems or taking initiative, at some point that begins to select for which employees stick around and get promoted. When people who embody that mentality become the workforce, or at least the management, it does get a little harder to say they're not at fault. Therefore, because a culture problem is at least partly a people problem, changing such a culture will necessarily involve breaking some eggs.

So, I'm not trying to let the execs, managers, or even rank-and-file employees off the hook. I'm just saying they're probably not the root of the problem and we just don't have enough information to know which of them really need to go. The one thing I'm certain about is that a lot of people are getting let go who were not bad apples, and it's just lying to make ourselves feel better to say otherwise.

I vehemently disagree and whole heartedly.so. Executives have one …. Job get I..t done. The whole point of a leader is to lead. Do you think people just listen because of position? Human beings have their own agendas, are highly predictable, yet, unpredictable, effective leaders are master manipulators period at navigating the human psyche to get what they want out of people . You can have all the best ideas and if you can neither convey them or implement them, you are not an effective leader. So absolutely I would blame such an ineffectual leader for the lack of strength, manipulation skills, savvy and/or ability to push their weight around and get ..it done. Period. The best executives certainly do not have a limitation mind set. No successful person has a limitation mindset.

 

[thestryker]

These cuts from the outside seem to be more of the downsizing/consolidation type than anything else. Pushing out people with experience before top to bottom industry standard nodes are running at IFS still doesn't seem like a great idea though.

Don't know about the others

Russell worked his way up from 2001 and was put in his current position by Gelsinger (process engineer to factory manager type positions).

Mistry less information easily available but came over from DEC so 1998 on. Not much on specific jobs within Intel though.

 

[TerryLaze]

Pushing out people with experience before top to bottom industry standard nodes are running at IFS still doesn't seem like a great idea though.

Article:

are set to retire,

I managed to find the age of one of them and he's 63.....
these are rich people, they don't need to keep working till they keel over.

Of course intel could try to keep them longer if they are really important and can't be replaced but they would have to find replacements pretty soon anyway.

 

[thestryker]

Article:

I managed to find the age of one of them and he's 63.....
these are rich people, they don't need to keep working till they keel over.

Of course intel could try to keep them longer if they are really important and can't be replaced but they would have to find replacements pretty soon anyway.

Sure just like Gelsinger retired.

I'm sure what's going on right now at Intel has nothing to do with it at all.

 

[bit_user]

I vehemently disagree and whole heartedly.so. Executives have one …. Job get I..t done. The whole point of a leader is to lead.

I'm just curious whether you have any experience working as a professional in a large company? Like, large enough to be listed on a stock exchange and have a few layers of middle managers.

Because, what I'm getting from your posts is that you have a sort of Hollywood impression of what a big company is like. Movies and TV are designed to be entertaining, not realistic. Having worked in tech for longer than I'd care to say, and for several different companies of various sizes, what you're saying just doesn't sound realistic to me.

 

[acadia11]

I'm just curious whether you have any experience working as a professional in a large company? Like, large enough to be listed on a stock exchange and have a few layers of middle managers.

Because, what I'm getting from your posts is that you have a sort of Hollywood impression of what a big company is like. Movies and TV are designed to be entertaining, not realistic. Having worked in tech for longer than I'd care to say, and for several different companies of various sizes, what you're saying just doesn't sound realistic to me.

The fact that you find the idea of effective leadership having the ability to actually impose their will, convey their ideas, and execute on their ideas to get the results they desire as a Hollywood idea is even more scary. And you’ve never been in the position to lead even at team level let alone an executive level. Literally an effective leader spends their entire day divining vision and then divining how to effect that vision and has no other purpose. I’m curious to know what you believe an effective leader does all day?

 

[abufrejoval]

Were these old executives good at their jobs? We know Intel foundries have not been doing the best for a while. It could be that these 3 were somewhat responsible for that.

Or they could have been great and will be dearly missed, or somewhere in the middle.

Without more information, judging whether this is good or bad news is premature.

To my understanding Intel never had a PDK before, because it was all about doing their own chips and their own tools for their own fab, not designed for abstraction and modularity but haphazardly put together for speed as the tick was tocking.

A PDK essentially allows you to translate your logical design to a physical process offered by a fab, so it's crucial for fabbing chips that you haven't designed yourself.

Building a PDK eco system isn't done in 7 days, but has been decades of experience with open fabs like TSMC or GF and even if you had that already in place you'd still have to extract the design rules that go into the PDK from Intel engineers.

I am an idiot at fabbing, but my gut feeling was that IFS is mission impossible, because you can't just catch up to a competition that has done these things for decades in years, especially because you also need to have the software tool vendors like Synopsis on-board, who put their efforts where the money is. And, of course, you need customers to trust and pay you.

No, Intel's choice to fast-track design to production in order to out-fab the competition was the recipe for their success for a long time. But it blew up in their face when out-fabbing stalled. They simply lacked the intermediate industry standard abstraction layers, but while the design teams managed to transition so that their chips could be fabbed externally, the effort for the fab turned out too big and too slow.

Intel is a giant and giants can't turn in days, months or even years. And no mattter how busy you are at the steering wheel, a course taken more than two decades ago, can't just be fixed when you finally see the iceberg.

Blaming won't help.

 

[usertests]

I'm wondering how many Intel employees AMD is picking up lately.

 

[abufrejoval]

I'm wondering how many Intel employees AMD is picking up lately.

AMD doesn't need those skills

 

[rluker5]

It sounds like Garry Patton was key to the IFS strategy and came along too recently to be blamed for execution failures on their process roadmap. Don't know about the others.

It's indeed possible that some house cleaning was necessary, in at least some of these cases. Getting rid of one of the guys key to the IFS strategy does indeed seem like a cause for concern, but I guess we won't know if it's going to affect IFS' business plan unless it does.

Some of these departures might've been on the initiative of the execs, who simply don't have faith in the company's direction. They're not all necessarily being fired or restructured out of jobs.

Tan apparently has disagreements with the scope of spending on fabs without enough confirmation of ROI. If Gary Patton was behind this it would definitely put him behind the 8 ball. This is just conjecture on my part as we don't have enough information, but it is certainly plausible that it is good, or at least neutral for Intel that Gary is gone. Not everyone good at some things is good at everything.

To my understanding Intel never had a PDK before, because it was all about doing their own chips and their own tools for their own fab, not designed for abstraction and modularity but haphazardly put together for speed as the tick was tocking.

A PDK essentially allows you to translate your logical design to a physical process offered by a fab, so it's crucial for fabbing chips that you haven't designed yourself.

Building a PDK eco system isn't done in 7 days, but has been decades of experience with open fabs like TSMC or GF and even if you had that already in place you'd still have to extract the design rules that go into the PDK from Intel engineers.

I am an idiot at fabbing, but my gut feeling was that IFS is mission impossible, because you can't just catch up to a competition that has done these things for decades in years, especially because you also need to have the software tool vendors like Synopsis on-board, who put their efforts where the money is. And, of course, you need customers to trust and pay you.

No, Intel's choice to fast-track design to production in order to out-fab the competition was the recipe for their success for a long time. But it blew up in their face when out-fabbing stalled. They simply lacked the intermediate industry standard abstraction layers, but while the design teams managed to transition so that their chips could be fabbed externally, the effort for the fab turned out too big and too slow.

Intel is a giant and giants can't turn in days, months or even years. And no mattter how busy you are at the steering wheel, a course taken more than two decades ago, can't just be fixed when you finally see the iceberg.

Blaming won't help.

I'm trying to stop blaming, conjecture and finger pointing.

But it doesn't seem as impossible for Intel to get into contract fabbing as you imply.
1. They are claiming their node is industry standard starting with 18A.
2. They claim their designs are now node agnostic.
3. They showed this with a last minute switch from Intel 20A to TSMC N3B with Arrow Lake when Intel 20A was abandoned. You can't keep your yearly cadence unless the design changes are known and small. That tape out stuff takes time. https://hardwaretimes.com/intel-15th-gen-arrow-lake-cpus-on-track-for-2nm-mass-production-in-2024/
4. I also don't know much about fabbing, but Synopsys is already on board, for decades even: https://www.synopsys.com/partners/intel-foundry-services.html and evidence of Intel's ability to contract out fabbing seems to be there. I even have an Intel fabbed x86 phone chip that was contract made for Spreadtrum in a Leagoo T5c: https://www.techradar.com/news/spreadtrum-unveils-high-performance-lte-platforms-sc9853i-and-sc9850

 

[abufrejoval]

I'm trying to stop blaming, conjecture and finger pointing.

But it doesn't seem as impossible for Intel to get into contract fabbing as you imply.
1. They are claiming their node is industry standard starting with 18A.
2. They claim their designs are now node agnostic.
3. They showed this with a last minute switch from Intel 20A to TSMC N3B with Arrow Lake when Intel 20A was abandoned. You can't keep your yearly cadence unless the design changes are known and small. That tape out stuff takes time. https://hardwaretimes.com/intel-15th-gen-arrow-lake-cpus-on-track-for-2nm-mass-production-in-2024/
4. I also don't know much about fabbing, but Synopsys is already on board, for decades even: https://www.synopsys.com/partners/intel-foundry-services.html and evidence of Intel's ability to contract out fabbing seems to be there. I even have an Intel fabbed x86 phone chip that was contract made for Spreadtrum in a Leagoo T5c: https://www.techradar.com/news/spreadtrum-unveils-high-performance-lte-platforms-sc9853i-and-sc9850

That sounds all good and fine, but

unfortunately, even with a PDK, customers can't just take their logical design and have a perfect physical design come out at the other end. The PDK is more a minimal pre-requisit, doesn't make a fab turn-key.

So customers have to trust Intel being able to sustain IFS over several generations at attractive speeds and prices in order to make their up-front investment; IFS would have had to build up a reputation.

And evidently potential customers weren't happy with the clues you list here. Well, according to Charlie D. some actually tried, and got bloody noses.

When AMD came out with Zen, it was clear that was a winner. But far fewer were convinced this wasn't a one-off. AMD knew that, and not only published a very clear roadmap three further generations out, but then executed on it, actually did even better than they predicted. And only with every next generation hitting the mark, the ranks of their customers swelled.

IFS needed ten years of success and they just didn't get them, too late and too little.

I'm sure China would love to buy, but everybody else might just pay for some scraps and the patents.

 

[rluker5]

That sounds all good and fine, but

unfortunately, even with a PDK, customers can't just take their logical design and have a perfect physical design come out at the other end. The PDK is more a minimal pre-requisit, doesn't make a fab turn-key.

So customers have to trust Intel being able to sustain IFS over several generations at attractive speeds and prices in order to make their up-front investment; IFS would have had to build up a reputation.

And evidently potential customers weren't happy with the clues you list here. Well, according to Charlie D. some actually tried, and got bloody noses.

I am not familiar with this Charlie D. And IFS 2.0 has had contracts go bad already? Like consoles at TSMC bad? (You couldn't get those for like a year because of lack of production.) News to me.

And Arrow Lake's transition seemed pretty turn-key.

 

[TerryLaze]

That sounds all good and fine, but

unfortunately, even with a PDK, customers can't just take their logical design and have a perfect physical design come out at the other end. The PDK is more a minimal pre-requisit, doesn't make a fab turn-key.

They know that, that's why they have programs for years now that help customers transition their designs over.
They also work together with arm and risc-v to be able to manufacture for a larger customer base.
https://www.intc.com/news-events/pr...undry-services-launches-ecosystem-alliance-to

Intel Foundry and Arm Announce Multigeneration Collaboration on Leading-Edge SoC Design

www.intc.com

Intel Corporation Makes Deep Investment in RISC-V Community to Accelerate Innovation in Open Computing – RISC-V International

riscv.org

So customers have to trust Intel being able to sustain IFS over several generations at attractive speeds and prices in order to make their up-front investment; IFS would have had to build up a reputation.

Even without the above, no they don't.
You can do your design node agnostic, as already mentioned, that way you can use intel or tsmc or whoever else, at the same time, or as a backup for each other.

When AMD came out with Zen, it was clear that was a winner. But far fewer were convinced this wasn't a one-off. AMD knew that, and not only published a very clear roadmap three further generations out, but then executed on it, actually did even better than they predicted. And only with every next generation hitting the mark, the ranks of their customers swelled.

First of all, what does that have to do with foundries?!?!?!
Second of all, if you think that CPU sales shows how good the FAB is then look at intel' revenue during that time.....

 

[abufrejoval]

I am not familiar with this Charlie D. And IFS 2.0 has had contracts go bad already? Like consoles at TSMC bad? (You couldn't get those for like a year because of lack of production.) News to me.

You may know him better as the guy behind semiaccurate.com. His material used to be quite good, but once he put things behind a paywall it's hard to tell. Also his sources at Intel might have dried up.

But there is a lot in the headlines already and it rhymes with outher sources, too.

And Arrow Lake's transition seemed pretty turn-key.

That's not an external customer fabbing at IFS...

 

[abufrejoval]

They know that, that's why they have programs for years now that help customers transition their designs over.
They also work together with arm and risc-v to be able to manufacture for a larger customer base.
https://www.intc.com/news-events/pr...undry-services-launches-ecosystem-alliance-to

Intel Foundry and Arm Announce Multigeneration Collaboration on Leading-Edge SoC Design

www.intc.com

Intel Corporation Makes Deep Investment in RISC-V Community to Accelerate Innovation in Open Computing – RISC-V International

riscv.org

Even without the above, no they don't.
You can do your design node agnostic, as already mentioned, that way you can use intel or tsmc or whoever else, at the same time, or as a backup for each other.

Ok, so why don't they have customers?

First of all, what does that have to do with foundries?!?!?!
Second of all, if you think that CPU sales shows how good the FAB is then look at intel' revenue during that time.....

It's about developing reputation so customers will trust you long term. And that doesn't happen with a single iteration, you need a track record of successes.

AMD executed on design for eight years and five generations and IFS is shooting blanks.