Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/intel-foundry-services-ushers-in-a-new-era.18081/
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021370
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

Intel Foundry Services Ushers in a New Era

Daniel Nenni

Admin
Staff member
First editorial from new IFS leader:

Intel Foundry Services Ushers in a New Era
Stuart Pann, Intel SVP and General Manager of Intel Foundry Services


New IFS leader shares his point of view on foundry progress, new collaborations, and opportunities for success.

In April, Intel Foundry Services (IFS) and Arm announced a multigenerational agreement to enable chip designers to build low-power compute system-on-chip (SoCs) on Intel technology. We are excited to provide our customers with the opportunity to design their mobile SoCs on Intel’s leading-edge 18A process technology paired with the latest, most powerful Arm CPU core — the recently launched next-generation Cortex-X4 — for improved power and performance.

This is a great sign of the momentum building around IFS, and it is part of the reason why I was eager to take on leadership of the organization in March. IFS is on a remarkable journey, and now that I’ve spent some time with the organization, I want to share more about why our mission is critically important for foundry customers around the world, for Intel and for me.

An Enormous Opportunity
I first joined Intel in 1981 at the dawn of the PC era and returned in 2021 when our CEO Pat Gelsinger introduced Intel’s IDM 2.0 strategy. In hearing his vision, I was motivated by the opportunity to help restore Intel’s global leadership in the growing semiconductor industry.

The global demand for semiconductors continues to see sustained, long-term growth, and the chip industry is expected to reach $1 trillion in sales by the end of the decade. This is an unprecedented opportunity. At the same time, 80% of the world’s chip manufacturing capacity is concentrated in Asia, and many foundry customers are looking for more options.

The industry needs resilient and globally diverse supply chains. Intel is one of only three companies in the world that currently make leading-edge chips. And until IFS was created, Intel was the only one without a commercial foundry. By tapping into Intel’s leading-edge manufacturing capabilities, legendary supply chain and strong partner ecosystem, IFS has an ambitious goal to become the second-largest foundry by 2030.

Our Differentiated Approach
Part of how we’ll power growth is by going beyond the traditional foundry offerings and building IFS as the world’s first open system foundry, leading the industry transition from standard monolithic system-on-chip to “systems of chips” in a package. Our combined offerings of wafer fabrication, advanced process and packaging technology, chiplet standards, software, robust ecosystem, and assembly and test capabilities will help our customers build innovative silicon designs and deliver full end-to-end customizable products.

Intel’s advanced packaging technologies are a key IFS differentiator, enabling customers to pack more features into each new generation while staying within the same cost, power and physical footprints. Today, some of our largest IFS customers, such as Amazon, Cisco and the U.S. Department of Defense (DoD), use our packaging solutions. The DoD needs leading-edge, onshore, advanced foundry and packaging capabilities more than ever, and IFS is ready to help the U.S. government make a secure transition to commercial foundries.

Customer-First
Intel’s people and culture have always been major pillars of our company’s success and resiliency – this was one of the reasons I returned. Over the past two years, Intel has developed key new skill sets and operating models to drive a cultural shift to a customer-first approach in support of foundry customers. This is a top priority for IFS as being a foundry means being a customer service business above all else and engaging deeply with customers throughout the design and manufacturing process.

As part of this shift within Intel, we treat external foundry customers and our internal business units equally, while providing confidentiality and protection for intellectual property (IP) and establishing dedicated customer capacity corridors. To do this, we’re implementing an internal foundry model that establishes consistent processes, systems and guardrails between our business units and our design and manufacturing teams. This puts external foundry customers on the same footing as our internal product groups. IFS customers can be assured of the technical excellence, innovation and capacity they’ve signed up for.

Implementing Our Strategy

We know from experience that building a successful foundry business quickly from the ground up is one of the hardest tasks we could take on. Now, in implementing our IDM 2.0 strategy, we’re taking steps to make sure we and our customers are successful.

Intel is preparing to support future demand from foundry customers through its global built-to-scale manufacturing capacity approach as it capitalizes on fab expansions in the U.S., EU and Israel. Getting our factories ready to support foundry customers will allow us to address the growing global demand for leading-edge chips, while also achieving the scale necessary to continue investing in advanced technologies.

We’re also building a robust ecosystem of best-in-class electronic design automation (EDA), silicon IP, design service, cloud, and U.S. military, aerospace and government alliances (USMAG) to help Intel’s foundry customers bring their silicon products from idea to implementation. Through the IFS Accelerator program, we deliver a seamless interface with Intel’s process technologies and accelerate customer innovation on IFS manufacturing platforms.

The Path Forward

IFS remains dedicated to ensuring that foundry customer products receive our utmost focus in terms of service, technology enablement and capacity commitments. As we build and expand on our future in foundry services offerings, we look to our strategic ecosystem partners, such as Arm, Cadence, Synopsys, and others, for support. We welcome the SoC design community to utilize the strength of Intel to drive the next level of semiconductor innovation.

IFS has come a long way in the past two years and I’m passionate about where we can take it in the years ahead. IFS will have its signature event later this year where we will talk about our progress and how we’re working with our ecosystem partners to power our customers’ innovations. I look forward to sharing more with you as we continue to implement our strategy, build out our capabilities and create a new leading-edge foundry alternative for the world.

Stuart Pann is Intel senior vice president and general manager of Intel Foundry Services.

 
Quickly penned response to the WSJ article?

THE WSJ article was old news. QCOM backed off from Samsung and Intel Foundry last year in favor of a closer relationship with TSMC. I have mentioned this before. This was not a "one PDK is better than the other" decision. QCOM has a new CEO so times are changing.

Just to be clear:

The big market for IFS is the NOT TSMC market which Samsung has enjoyed for the last few years. Everyone wants a second or third choice for the leading edge foundry business (including myself) so that is where we are at today. Unfortunately the gap between TSMC and Samsung in regards to trusted PDKs is widening. If Intel can produce a strong 18A foundry PDK they will win that business, absolutely.

The MADE IN USA foundry business was also in favor of Intel but now TSMC is building in AZ so that is in question. I still feel Intel still has the advantage but it could be argued both ways.

In order to win the BETTER THAN TSMC business IFS has a longer way to go. From what I know today IFS 18A and TSMC N2 are competitive processes in regards to technology and HVM roadmaps. IFS and TSMC packaging are also competitive. Intel can claim firsts in some areas due to their internal product work but TSMC holds most of the packaging firsts for the foundry business. Once Intel ramps up their foundry customers this may change. The TSMC ecosystem is still far superior but IFS is working on that and can easily support some big customers as they mostly use internal IP and have the full support of EDA vendors. Either company can still fumble so the 18A/N2 game is not over yet.

The latest move that IFS made, which I am a fan of, is the change in leadership. Stuart Pann is the right guy to lead this new foundry effort. From what I have been told he is a strong no BS leader and has the trust of the CEO and board. He also listens to his team and takes action. I know some of his team and they are the right people to do this job so as long as the Intel process technology holds up in the foundry world they will succeed.

The only downside I see is expectations. The foundry business is a marathon not a sprint and if you take short cuts you will be disqualified. It will take years for a foundry to jump from the NOT TSMC market to the BETTER THAN TSMC market so be patient.

Just my opinion of course but according to chat.openai.com I'm a respected figure in the semiconductor community. :cool:
 
Quickly penned response to the WSJ article?


About Mr. Stuart Pann:

"Stuart Pann is senior vice president and general manager of Intel Foundry Services (IFS). In this role, Pann drives continued growth for IFS and its differentiated systems foundry offering, which goes beyond traditional wafer fabrication to include packaging, chiplet standards and software, as well as U.S.- and Europe-based capacity.

Pann previously served as chief business transformation officer and general manager of Intel’s Corporate Planning Group. As part of this role, he established the company’s IDM 2.0 Acceleration Office (IAO) to guide the implementation of an internal foundry model. IAO closely collaborates with all Intel business units and functional teams to support the company’s internal foundry model.

In June 2021, Pann returned to Intel, where he had started his career in 1981. Prior to his return, he was chief supply chain officer and chief information officer at HP for six years. At HP, Pann was responsible for the company’s supply chain, which delivers nearly 100 million products to customers each year.

Before joining HP in July 2014, Pann served as corporate vice president and general manager of Intel’s Business Management Group, where he was responsible for pricing, revenue and forecasting functions for the company’s microprocessor and chipset operations. He also co-managed the geographic operations teams for the Intel sales force and was responsible for order management and external-facing supply chain programs. Pann held several management positions within the company’s sales organization before moving into an operations role in 1999 as the director of Microprocessor Marketing and Business Planning.

Pann earned a bachelor’s degree in electrical engineering from Michigan Technological University and an MBA from the University of Michigan."

Source: https://www.intel.com/content/www/us/en/newsroom/biographies/biography-stuart-pann.html#gs.z5zyto:
 
THE WSJ article was old news. QCOM backed off from Samsung and Intel Foundry last year in favor of a closer relationship with TSMC. I have mentioned this before. This was not a "one PDK is better than the other" decision. QCOM has a new CEO so times are changing.

Just to be clear:

The big market for IFS is the NOT TSMC market which Samsung has enjoyed for the last few years. Everyone wants a second or third choice for the leading edge foundry business (including myself) so that is where we are at today. Unfortunately the gap between TSMC and Samsung in regards to trusted PDKs is widening. If Intel can produce a strong 18A foundry PDK they will win that business, absolutely.

The MADE IN USA foundry business was also in favor of Intel but now TSMC is building in AZ so that is in question. I still feel Intel still has the advantage but it could be argued both ways.

In order to win the BETTER THAN TSMC business IFS has a longer way to go. From what I know today IFS 18A and TSMC N2 are competitive processes in regards to technology and HVM roadmaps. IFS and TSMC packaging are also competitive. Intel can claim firsts in some areas due to their internal product work but TSMC holds most of the packaging firsts for the foundry business. Once Intel ramps up their foundry customers this may change. The TSMC ecosystem is still far superior but IFS is working on that and can easily support some big customers as they mostly use internal IP and have the full support of EDA vendors. Either company can still fumble so the 18A/N2 game is not over yet.

The latest move that IFS made, which I am a fan of, is the change in leadership. Stuart Pann is the right guy to lead this new foundry effort. From what I have been told he is a strong no BS leader and has the trust of the CEO and board. He also listens to his team and takes action. I know some of his team and they are the right people to do this job so as long as the Intel process technology holds up in the foundry world they will succeed.

The only downside I see is expectations. The foundry business is a marathon not a sprint and if you take short cuts you will be disqualified. It will take years for a foundry to jump from the NOT TSMC market to the BETTER THAN TSMC market so be patient.

Just my opinion of course but according to chat.openai.com I'm a respected figure in the semiconductor community. :cool:

"The latest move that IFS made, which I am a fan of, is the change in leadership. Stuart Pann is the right guy to lead this new foundry effort. From what I have been told he is a strong no BS leader and has the trust of the CEO and board. "

It seems that Stuart Pann doesn't have firsthand experience in leading/managing a foundry business. I don't understand why Intel can't recruit one with strong foundry background to lead IFS?

The time widow for IFS to be successful is very short due to the constraints of market, technology changes, and Intel's own financial standing. Intel is spending billions of money to build IFS then why it wants to let someone lead IFS with no prior experience?
 
Last edited:
"The latest move that IFS made, which I am a fan of, is the change in leadership. Stuart Pann is the right guy to lead this new foundry effort. From what I have been told he is a strong no BS leader and has the trust of the CEO and board. "

It seems that Stuart Pann doesn't have firsthand foundry business experience. I don't understand why Intel can't recruit one with strong foundry experience to lead IFS?

The time widow for IFS to be successful is very short due to the constraints of market, technology changes, and Intel's own financial standing. Intel is spending billions of money to build IFS then why it wants to let someone lead IFS with no prior experience?

I disagree. A person with experience but without a close relationship with Pat G is of no use. Big decisions will be required. Stuart is surrounded by outsiders who are very foundry experienced. Bottom line: The IFS team knows what to do but they need an Intel trusted leader to get it done and yes time is of the essence. In my opinion Intel is all in on the IDM 2.0 strategy and the foundry business is a big part of it.
 
I disagree. A person with experience but without a close relationship with Pat G is of no use. Big decisions will be required. Stuart is surrounded by outsiders who are very foundry experienced. Bottom line: The IFS team knows what to do but they need an Intel trusted leader to get it done and yes time is of the essence. In my opinion Intel is all in on the IDM 2.0 strategy and the foundry business is a big part of it.

I'm not arguing that the IFS needs a leader who is trusted by the Intel Board of Directors and CEO Pat Gelsinger. It is a must.

My question is that why Intel can't find a leader for this billion dollar IFS venture who is not only trusted but also has strong foundry experience? Is it because there is no such person existing on the earth at all?

Looking back Intel's recent history, it had lots of people in the senior leadership positions who were trusted by Intel's then Boards and CEOs. Most of them are nice people but many of them were assigned to the wrong positions with mismatched skills, experiences, and expectations.
 
I'm not arguing that the IFS needs a leader who is trusted by the Intel Board of Directors and CEO Pat Gelsinger. It is a must.

My question is that why Intel can't find a leader for this billion dollar IFS venture who is not only trusted but also has strong foundry experience? Is it because there is no such person existing on the earth at all?

Looking back Intel's recent history, it had lots of people in the senior leadership positions who were trusted by Intel's then Boards and CEOs. Most of them are nice people but many of them were assigned to the wrong positions with mismatched skills, experiences, and expectations.

Unfortunately CC Wei was not available. :) Other than CC who would you chose?

If the Tower acquisition had gone through the Tower CEO would probably would be running IFS as an independent business unit but that did not happen and probably won't from the looks of it.

As you said time is of the essence so a person with Stuart's creds is the safe pick.
 
Unfortunately CC Wei was not available. :) Other than CC who would you chose?

If the Tower acquisition had gone through the Tower CEO would probably would be running IFS as an independent business unit but that did not happen and probably won't from the looks of it.

As you said time is of the essence so a person with Stuart's creds is the safe pick.

TSMC's CC Wei and Tower Semi's Russell Ellwanger are good but shouldn't be the only two in the whole world. It's a good topic to start a new thread. Although Mr. Stuart Pann probably won't be happy about it 🙂🙄.
 
I'm not arguing that the IFS needs a leader who is trusted by the Intel Board of Directors and CEO Pat Gelsinger. It is a must.

My question is that why Intel can't find a leader for this billion dollar IFS venture who is not only trusted but also has strong foundry experience? Is it because there is no such person existing on the earth at all?

Looking back Intel's recent history, it had lots of people in the senior leadership positions who were trusted by Intel's then Boards and CEOs. Most of them are nice people but many of them were assigned to the wrong positions with mismatched skills, experiences, and expectations.
Look at the CEO that trusted them. Gelsinger is, in my opinion, far better than that swine Krzanich or Bob Swan. Krzanich showed poor judgment repeatedly, one such instance lead to his ouster.

Certain things you can learn, some things you can't. I'd rather have someone that is very capable, but not as experienced, than someone less capable, but more experienced, for that reason. Experience can be gained, but if this guy has all the other boxes checked, or is better than other candidates in any case, I can see why they chose him, and surrounded him with people that can help mitigate his lack of experience.

I'm not sure about this time is of the essence stuff either. Sure, you always want to achieve success as quickly as possible, but even if Intel isn't widely successful right off the bat, they still have a captive customer if need be (themselves), and they still need to invest in process technology. I'm not sure how it's a marathon, but also "time is of the essence". They'll be around in a few years, given the market they are trying to play in, so will companies looking eagerly for them to offer a compelling solution. I think the marathon analogy is dead on.

But, no doubt it's better if they get things together sooner than later, but it's not exactly a race to the wall either. As a big Intel stockholder, I'm fine if they keep making relative improvements to their market positioning, and gain momentum and competence in a linear way. My fear isn't so much timing, but will they continue to offer compelling scenarios, and will they continue to improve as they plow forward. If so, I think they'll be OK, if not, they'll lose the confidence of their customers, and that's a really hard thing to gain back. But, if NVIDIA says, well, we really wanted to use 18A, but it's not quite ready, but your next one we can use, I think that's a good scenario for Intel.

I think to some extent we're already seeing something like that. Intel was talking pretty boldly about Intel 3 at one point, but that sure seems tempered at this point. By all indications I have seen, and I could be wrong, they're going to get some wins there, but it's much more heavily weighted towards 18A than it was. And I think it could be playing out as I talked about in the prior paragraph. But, I don't really know, that's for sure.

I do believe, however, getting the best person is a good idea, rather than the most experienced, because I think it has better long term implications. Of course, I assume that they felt he was that person, and that's why they put him in that role.
 
An alternate take on the Stuart Pann debate here. The two perspectives shared so far are a) trust with Pat / Intel Board is important; b) foundry experience is important. The third I would add is c) trust with potential customers is important. This in a way leads back to the foundry experience point - foundries tend to have a customer centric culture which Intel seems to lack. This is what cements customer trust. However, this customer centiricity is not unique to foundries and can be found in a wide range of businesses. If Intel doesnt build this customer centricity, it would end up like Samsung - second source to TSMC, even if they are competitive on technology.
 
An alternate take on the Stuart Pann debate here. The two perspectives shared so far are a) trust with Pat / Intel Board is important; b) foundry experience is important. The third I would add is c) trust with potential customers is important. This in a way leads back to the foundry experience point - foundries tend to have a customer centric culture which Intel seems to lack. This is what cements customer trust. However, this customer centiricity is not unique to foundries and can be found in a wide range of businesses. If Intel doesnt build this customer centricity, it would end up like Samsung - second source to TSMC, even if they are competitive on technology.
I spent many years working in the ASIC business and got to think of it as one where you had to be good - let's say at least 80% - at a wide and diverse range of activities. Being best (100%) at only two or three usually wasn't enough if several others were weak. I suspect foundry is similar and that this is one of Intel's big challenges in moving from being a culture where excellence in two or three things was all that mattered and they so dominated the markets they served that customers got used to having to listen to them rather than the other way round.

I view the customer centricity as essential to even be in the foundry business.

Intel are now saying all the right things about this. Too early to judge how this is going to work out in practice.
 
THE WSJ article was old news. QCOM backed off from Samsung and Intel Foundry last year in favor of a closer relationship with TSMC. I have mentioned this before. This was not a "one PDK is better than the other" decision. QCOM has a new CEO so times are changing.

Just to be clear:

The big market for IFS is the NOT TSMC market which Samsung has enjoyed for the last few years. Everyone wants a second or third choice for the leading edge foundry business (including myself) so that is where we are at today. Unfortunately the gap between TSMC and Samsung in regards to trusted PDKs is widening. If Intel can produce a strong 18A foundry PDK they will win that business, absolutely.

The MADE IN USA foundry business was also in favor of Intel but now TSMC is building in AZ so that is in question. I still feel Intel still has the advantage but it could be argued both ways.

In order to win the BETTER THAN TSMC business IFS has a longer way to go. From what I know today IFS 18A and TSMC N2 are competitive processes in regards to technology and HVM roadmaps. IFS and TSMC packaging are also competitive. Intel can claim firsts in some areas due to their internal product work but TSMC holds most of the packaging firsts for the foundry business. Once Intel ramps up their foundry customers this may change. The TSMC ecosystem is still far superior but IFS is working on that and can easily support some big customers as they mostly use internal IP and have the full support of EDA vendors. Either company can still fumble so the 18A/N2 game is not over yet.

The latest move that IFS made, which I am a fan of, is the change in leadership. Stuart Pann is the right guy to lead this new foundry effort. From what I have been told he is a strong no BS leader and has the trust of the CEO and board. He also listens to his team and takes action. I know some of his team and they are the right people to do this job so as long as the Intel process technology holds up in the foundry world they will succeed.

The only downside I see is expectations. The foundry business is a marathon not a sprint and if you take short cuts you will be disqualified. It will take years for a foundry to jump from the NOT TSMC market to the BETTER THAN TSMC market so be patient.

Just my opinion of course but according to chat.openai.com I'm a respected figure in the semiconductor community. :cool:

I think the biggest market for IFS is always Intel itself between now - 2030. IFS needs to show that it can beat TSMC in terms of performance, efficiency, and cost. Of course, it's extremely hard for them to do all three elements, but it is now progressing well, and ahead of TSMC roadmap when it comes to 18A. What we are now seeing in Intel's upcoming products portfolio. Meteor Lake, Arrow Lake, Alchemist/+, Battlemage, etc, will be made exclusively or majority of the chiplets at TSMC. That is not a good position to be at considering the fact that IDM makes a greater profit margin through vertical integration.

I hold a different than Daniel that it's not about "NOT TSMC", "MADE IN AMERICA", or "BETTER THAN TSMC" and the order of these three. I believe the best market for IFS is always performance-first, TCO first workloads. Specifically, data center + custom chips designed by big cloud players. Most notably, AWS, GCP, and Azure. These cloud hyperscalers have massive and outdated x86 servers waiting to be replaced. Will they go to AMD for Ryzen or continue with Xeon? maybe both or either one. But I bet that they will more than likely to create their own customized x86 servers with the helps from Intel and some other IP companies because that's what determining their competitiveness after watching what AWS did with their homegrown chips. Even if they don't choose x86 as the instruction set. They can still have a lower total cost of ownership through the most leading-edge node, and that is 18A.

Another thing that's really important is time to market. AMD beats Intel in recent years because it is using TSMC, same with ARM in data centers, and Apple. And they come to market a lot earlier than Intel do with Intel 7. With the launch of Seirra Forest in 2024 on Intel 3, with AMD following suit with N3E, they will be neck & neck. And then Clearwater Forest in 18A in 2025, believe it or not. If Intel succeed in launching 18A products, I see plenty of chip designers (specifically data centers) to switch to Intel because they have to. TSMC will only be able to ramp up production of N2, and N2P at 2026~2027, with Apple being the first to market at 2025. This means that they can only catch up to Intel 1~2 years later. I don't think that's a good thing for AMD, for Nvidia, and for Ampere because they are late and the best chips will already be in market 1 year before they do.

To summarize, the best market to attack is the adjacent markets of x86, specifically data center because majority of the market is dominated by x86. And the key ingredient to win is time to market. (at this stage)
 
I think the biggest market for IFS is always Intel itself between now - 2030. IFS needs to show that it can beat TSMC in terms of performance, efficiency, and cost. Of course, it's extremely hard for them to do all three elements, but it is now progressing well, and ahead of TSMC roadmap when it comes to 18A. What we are now seeing in Intel's upcoming products portfolio. Meteor Lake, Arrow Lake, Alchemist/+, Battlemage, etc, will be made exclusively or majority of the chiplets at TSMC. That is not a good position to be at considering the fact that IDM makes a greater profit margin through vertical integration.

I hold a different than Daniel that it's not about "NOT TSMC", "MADE IN AMERICA", or "BETTER THAN TSMC" and the order of these three. I believe the best market for IFS is always performance-first, TCO first workloads. Specifically, data center + custom chips designed by big cloud players. Most notably, AWS, GCP, and Azure. These cloud hyperscalers have massive and outdated x86 servers waiting to be replaced. Will they go to AMD for Ryzen or continue with Xeon? maybe both or either one. But I bet that they will more than likely to create their own customized x86 servers with the helps from Intel and some other IP companies because that's what determining their competitiveness after watching what AWS did with their homegrown chips. Even if they don't choose x86 as the instruction set. They can still have a lower total cost of ownership through the most leading-edge node, and that is 18A.

Another thing that's really important is time to market. AMD beats Intel in recent years because it is using TSMC, same with ARM in data centers, and Apple. And they come to market a lot earlier than Intel do with Intel 7. With the launch of Seirra Forest in 2024 on Intel 3, with AMD following suit with N3E, they will be neck & neck. And then Clearwater Forest in 18A in 2025, believe it or not. If Intel succeed in launching 18A products, I see plenty of chip designers (specifically data centers) to switch to Intel because they have to. TSMC will only be able to ramp up production of N2, and N2P at 2026~2027, with Apple being the first to market at 2025. This means that they can only catch up to Intel 1~2 years later. I don't think that's a good thing for AMD, for Nvidia, and for Ampere because they are late and the best chips will already be in market 1 year before they do.

To summarize, the best market to attack is the adjacent markets of x86, specifically data center because majority of the market is dominated by x86. And the key ingredient to win is time to market. (at this stage)
Hyper scalers aren’t going to switch back and forth from Intel to Amd just like that. Amd has permanently taken market share. 18A being a leader is still very much in doubt as well. Many seem eager to arrest crown it the next best process. I doubt N2+ won’t be at least as good if not better…
 
18A being a leader is still very much in doubt as well. Many seem eager to arrest crown it the next best process.
Based on the current roadmaps we can all see right now, 18A will be the most advanced node and offer the best performance@iso power when it is planed to come out. As for people being quick to award it that crown I would say just as many folks online (folks I would argue have little understanding on this topic) seem to think it won’t even be competitive with N3E or that it won’t come out until 2028. But you are certainly right that it is not ordained by providence for 18A to be an unquestioned leader on PPW and competitive with N2 on all other metrics. However 18A failing and N2 succeeding or failing are also not predetermined outcomes. At the end of the day intel and TSMC engineers are going to have to put their money where their mouths are and prove it. I for one expect great things from both of them :giggle:

I doubt N2+ won’t be at least as good if not better…
Okay… so? 10FF was more dense than intel 14nm, but launched iPhones 2-3 years after skylake/broadwell, and when it came out 14nm was hands down the most advanced node on the market by every metric. Is it really so bad for intel if 18A has worse density and comparable PPW to N2P, if intel starts ramping 18A in 2H24 1.5-2 years before N2P (depending on if it ramps during 1H or 2H26)? Let me reiterate, this is NOT me prophesying “TSMC is doomed!”. Rather my point is that IF intel delivers to it’s roadmap, I think that is without a doubt a return to form for the former technological leader of the semiconductor manufacturing industry. Intel orchestrating a technological comeback is not in my opinion something that is mutually exclusive with TSMC’s continuing to be successful.
 
An alternate take on the Stuart Pann debate here. The two perspectives shared so far are a) trust with Pat / Intel Board is important; b) foundry experience is important. The third I would add is c) trust with potential customers is important. This in a way leads back to the foundry experience point - foundries tend to have a customer centric culture which Intel seems to lack. This is what cements customer trust. However, this customer centiricity is not unique to foundries and can be found in a wide range of businesses. If Intel doesn't build this customer centricity, it would end up like Samsung - second source to TSMC, even if they are competitive on technology.

IFS hired some key TSMC and Samsung people who are close to customers. Those are the guys that will get the job done in the trenches. At the executive level Stuart is the right guy as long as he enables the people in the trenches. My guess is that, after listening to his people, Stuart will reorg and hit the customer base aggressively with a new strategy. At least that is what I would do. Intel has an investor foundry call next week, we should know more then. Exciting times in the foundry business, absolutely.
 
IFS hired some key TSMC and Samsung people who are close to customers. Those are the guys that will get the job done in the trenches. At the executive level Stuart is the right guy as long as he enables the people in the trenches. My guess is that, after listening to his people, Stuart will reorg and hit the customer base aggressively with a new strategy. At least that is what I would do. Intel has an investor foundry call next week, we should know more then. Exciting times in the foundry business, absolutely.

An alternate take on the Stuart Pann debate here. The two perspectives shared so far are a) trust with Pat / Intel Board is important; b) foundry experience is important. The third I would add is c) trust with potential customers is important. This in a way leads back to the foundry experience point - foundries tend to have a customer centric culture which Intel seems to lack. This is what cements customer trust. However, this customer centiricity is not unique to foundries and can be found in a wide range of businesses. If Intel doesnt build this customer centricity, it would end up like Samsung - second source to TSMC, even if they are competitive on technology.

It also displayed two different leadership approaches. We can see several new Intel key executives who worked for/with Pat Gelsinger during Pat's time in Intel and VMware. Pat brought people he knew and trusted to his inner circle to rebuild Intel.

About two years ago during a speech in Taiwan, Morris Chang recalled an advice his old Texas Instruments acquaintance gave to him when he just started TSMC. That person suggested that Morris should build his leadership team with people he knew, like those who have worked for TI when Morris was there. That person insisted otherwise it could lead to a tragic failure.

Morris Chang strongly disagreed with that. Morris mentioned that his typical style when he joined a new company or a new division is to bring no one from his past to the new organization. At most he only brought his personal secretary with him to the new position.

Morris explained he believes to build a team with people he knew before will create barrier to the trust and cooperation and limit the fresh ideas, vision, and innovations. Worst it can create multiple factions inside an organization that will hurt the company greatly.

Obviously both approaches have their pros and cons under different situations.
 
Last edited:
It also displayed two different leadership approaches. We can see several new Intel key executives who worked for/with Pat Gelsinger during Pat's time in Intel and VMware. Pat brought people he knew and trusted to his inner circle to rebuild Intel.

About two years ago during a speech in Taiwan, Morris Chang recalled an advice his old Texas Instruments acquaintance gave to him when he just started TSMC. That person suggested that Morris should build his leadership team with people he knew, like those who have worked for TI when Morris was there. That person insisted otherwise it could lead to a tragic failure.

Morris Chang strongly disagreed with that. Morris mentioned that his typical style when he joined a new company or a new division is to bring no one from his past to the new organization. At most he only brought his personal secretary with him to the new position.

Morris explained he believes to build a team with people he knew before will create barrier to the trust and cooperation and limit the fresh ideas, vision, and innovations. Worst it can create multiple factions inside an organization that will hurt the company greatly.

Obviously both approaches have their pros and cons under different situations.

One approach is for starting a new company and the other is for pivoting an existing business? Someone said turning Intel is like turning the Titanic, hopefully avoiding the iceberg this time. You definitely want trusted hands for that one.
 
One approach is for starting a new company and the other is for pivoting an existing business? Someone said turning Intel is like turning the Titanic, hopefully avoiding the iceberg this time. You definitely want trusted hands for that one.

It may be true that Morris' approach is more suitable for starting a new company but he did it anyway even when he took an executive job in a different division of a company he already worked for or for a long existing company/organization he didn't work before. There are many personal and organization's factors that can affect an executive's decision to choose one of the best approaches he/she believes.

In Pat Gelsinger's vision, is ISF a brand new business (almost like a new startup) or an IDM happened to add the foundry service to its official line of product and service? It will have very different implications.
 
It may be true that Morris' approach is more suitable for starting a new company but he did it anyway even when he took an executive job in a different division of a company he already worked for or for a long existing company/organization he didn't work before. There are many personal and organization's factors that can affect an executive's decision to choose one of the best approaches he/she believes.

In Pat Gelsinger's vision, is ISF a brand new business (almost like a new startup) or an IDM happened to add the foundry service to its official line of product and service? It will have very different implications.
I would definitely view IFS as a brand new business rather than pivoting an existing business. Perhaps ultimately a quite separate business from Intel.

I'm with Morris Chang's view here. TI was notorious for not hiring from outside. And also - in the 1990s - for a slide showing their slow, monotonic decline from #1 in semis to the bottom half of the top 10. Not a coincidence in my view. Not meaning to have a go at TI there - a great company and I spent many years there and am pleased they recovered.
 
Back
Top