So, my name is Toshiya Hari. I cover the semiconductor space at Goldman Sachs. We're very pleased, very honored to have Stu Pann from Intel with us this morning, still, I think.
Yes.
Just a brief intro. Stu is Senior Vice President and General Manager of Intel Foundry Services, AKA IFS. In this role, he 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 E.U.-based capacity. Stu 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 to guide the implementation of an internal foundry model. Before handing the mic over to Stu, please note that today's discussion may contain forward-looking statements that are subject to various risks and uncertainties and may reference non-GAAP financial measures.
Please refer to Intel's most recent earnings release and annual report in Form 10-K and other filings with the SEC for more information on the risk factors that could cause actual results to differ materially and additional information on our non-GAAP financial measures, including reconciliations where appropriate, to the corresponding GAAP financial measures. With that, Stu w ell o ver to you.
Thank you. Well, given the audience, I, I've noticed that you guys probably aren't wanting to listen to a lot of slides, so I'm actually gonna breeze through these things really quickly, and then we're just gonna turn it over for conversation. If you look at what's going on behind our transformation, I first and foremost, you know, Intel is a manufacturing company. We're focused on process technology. You've heard my boss, Pat Gelsinger, talk any number of times about 5 nodes in 4 years. We are on track to get to 5 nodes in 4 years, and only by having process technology equivalency and then superiority, do you actually make margin in this business.
So we are focused on that, and I'm pleased to report we're making good progress across all of our different nodes, where the stuff we're ramping today, we talked last week about Meteor Lake, now in volume production, for product launches at the end of the year. Granite Rapids, our server part, Intel 3, is taping out and sampling now, and we're now working on 20 and 18A designs. So I think good progress on the execution front. On my next slide, I'll talk a little bit more about what the foundry has in terms of business offering. And the third element, which I think is really important as people look at the company overall, is what role does Intel play in AI?
We play probably a, not a well-known, but I think a significant role in AI because we have AI characteristics embedded in our server processors and embedded in our mobile processors, and you'll see a lot of that play out over the coming several months as we announce more and more capability there. I'd rather spend time on really the foundry business, and I have this little, triangle, if you will, this pyramid of capability. To be successful in the foundry business, first and foremost, there are table stakes. You have to have competitive cost, you have to have competitive power, you have to have competitive scaling, and you have to have a stable process roadmap. Without that, you can't play in this business.
We have learned a lot over the last couple of years of how to go about doing that, and 18A, which is our leading-edge node, is the first process that we've designed from the ground up to have all those table stakes. This is why we had the analysis with Synopsys a few weeks ago. This is where we're filling out all the IP necessary to make this a reality, but it's not just going to be about the IP, it's about having all the tools, flows, and methods necessary to create a device, and we call this dial tone. When a customer calls up, they want to know that from start to finish, as they're working on us with designs, we can take it all the way through to production. So we're filling out all the portfolio to have dial tone, and that's the big first step.
But on top of that, we can add other assets to the company. We're clearly getting a lot of interest in packaging today. As you probably have heard, there's a little bit of a shortage of packaging in the industry, and we have a lot of capabilities in this area because we've been in the server business for so long. So a lot of inquiries from customers around the world saying, "How can you help us with more kinds of packaging capability?" But then you build on even more on that, and there are things that we bring to the party in terms of systems and platform expertise.
Because we're in the server business, because we're in the desktop business, because we know a fair amount about software, we're able to help our customers at a new, a unique level because we understand what they're trying to build, so we're out trying to monetize that software and services and platform knowledge in order to create competitive advantage. And that's simply how we think about the business is table stakes, ingredients like packaging, and then capabilities to create what we call a systems foundry. And that's the strategy in a nutshell, but I'll turn it over for our questions and conversation with all of you for the next 30-some minutes or so.
That's great. Thank you so much for the overview. Before we dive into the business, wanted to ask a question more about you. You started your career at Intel about 40 years ago.
Forty-two.
42 years ago. You left for HP in 2014.
I did.
And then you returned to Intel in 2021.
What catalyzed the move to HP? What catalyzed the move back to Intel? And what was your first reaction, impression, when you were asked to lead Yea the foundry business?
, it was interesting. You know, I left in 2014. I was running operations for Paul Otellini from 2000 to 2013, and then Paul, as many of you know, retired in mid-2013. I spent a year working for Brian Krzanich. He had one vision of the company, which included things like drones and watches and clothes and things a little removed from our core mission, I thought. So he was the CEO. I was not, so I decided to go work for Dion Weisler and Meg Whitman at HP for 6 years as the Chief Supply Chain Officer, and also as the CIO of the company, and had a great time, learned a lot. We created a lot of shareholder value with the team.
But I also got a sense of how Intel strategies played out in the marketplace, and it was fascinating to see this, you know, whole circle of life play out. I retired from HP shortly after Dion retired, did a startup in the middle of COVID for three months. That was a really bad idea. Retired again, and then started doing a little bit of consulting for Intel, and then it was really Pat that brought me back and Pat's vision for the company. And so I've worked with him on rebuilding a lot of the core planning processes we had in the company that allowed us to hit, you know, tick-tock, this idea that you have process cadence and product cadence. So we worked on that for a couple of years.
I think it's now starting to bear fruit, is the fact that we're hitting our milestones, our products are either stable on their schedule or even pulling in a little bit. And then back in March, Pat called me up at 6:00 A.M. to ask me if I wanted to run the foundry business, and my reaction was: I'm really glad my wife is sleeping right now. Because she, she had thought at some point I was gonna retire. And I decided not to, and so I decided to sign up for another tour of duty and help Pat build out this capability, because being in the foundry business is essential for Intel. It gives us scale, it gives us knowledge, it gives us compatibility.
And, you know, I think as the more I get into this role, I've been now doing it for about five and a half months, just the learning we have from talking to customers as foundry partners, the insights we're getting is gonna make Intel a stronger company. And Pat's fond of saying, "Our foundry business makes our IDM business, our server and, and client business, more competitive," and vice versa, and that's absolutely turned out to be the case.
That's interesting. As the head of IFS, you know, how do you spend your time? What are your near-term priorities, long-term goals, and how are you evaluated as the head of IFS, which obviously in the near term, it's still a relatively small percentage of the business.
It is.
It's expected to grow over time. Pat and the board, how do they evaluate you?
, I mean, job one this year is to go land Intel 18A customers. And, for those of you in the audience who wanna ask me who the prepaid customer was, I'll tell you ahead of time, I'll probably disappoint you in my answer, but you can ask it if you want. But the job really is to go land, you know, serious customers that can drive significant volumes. We, you know, we have landed and working closely with MediaTek on Intel 16. We have landed a compute storage partner in Intel 3. We have landed at least a prepay commitment, but we obviously would like to be able to make more definitive statements by the end of the year on this.
So the bulk of my time is really working with customers and making sure we're lined up to service their needs and making sure they're comfortable with us as with our capabilities as a foundry supplier. I'd say about 60-70% of my time is based on that. Another 20% of my time is really based on building the organization. Do we have the right talent in place? A lot of my staff actually have come from foundry partners because we want to build that DNA inside of the organization. So we're looking at how do we recruit, how do we bring talent in, how do we make sure we've got a cohesive framework so they feel they can contribute right out the chute.
And then probably 10, 10% of the time is, you know, spent, you know, really on some of the long-range planning stuff, but we're really focused on proving to the world we can land customers today. And boy, part of those customers include working closely with the U.S. government on the CHIPS Act funding and also working with, you know, helping with the EU proposals that we've made, because they are our customers, right? They want to have that same level of confidence that as they're giving us taxpayer dollars, that we will be a wise steward of them and make that all pay off. So I would put, you know, governments inside of that 70% bucket.
Great. I guess in your prepared remarks, you talked about table stakes, you talked about the ingredients.
.
In your view, what makes a great, strong foundry supplier?
-
Which aspects do you possess today, and which aspects do you still need to work on?
Well, I think what, you know, in terms of what we have to, you know, coming out of the chute, a lot of people don't realize this, but I have to, and I spent time pointing it out, is, you know, we have a really good IDM business. You know, our Q2 guidance was $12.9 billion-$13.9 billion. Pat last week guided to the higher end of that range. But because we have the assets of a strong internal business, we've 40,000 people in manufacturing, 15,000 people in R&D, and we can leverage off of that base. The challenge for us is how do we take that base that's been internally focused and make it externally focused?
And I think as we show the world that we have process capability, like we do in, you know, in 18A and 20A, and, you know, Intel 7, 4, and 3, as we show the world we can deliver to, to demanding packaging requirements, which are all coming out of Singapore, I think those are things we can build out, but we have a lot to work from. You know, we're doing, in terms of Moore's Law, Pat is fond of saying that Moore's Law will be exhausted in the last element of the periodic table, so we have a ways to go there. But clearly, for the next 10 years, we can keep knocking out process improvements year after year. So that's our base, that's our heritage.
Now, the question is: how do we take that heritage and make it available for all foundry customers and not just the Intel internal business units? And I think, , we're making headway to that. To that end, I think some of you may have heard about the fact that we're doing an internal operating model for the foundry business and for our business units. And that's one of the key points that we're working on really hard, is how do you separate manufacturing and foundry from our product groups? We're creating a structure that involves arm's-length transactions between manufacturing and the Intel internal business units, so they can buy at a wafer cost, TSMC-like... so they can get defect densities at a TSMC-like defect density, and it's incumbent on the organization that does manufacturing to hit those targets.
So now our business units can plan their business the same way AMD plans their business with TSMC or Qualcomm does, and then we're very focused on hitting our cost targets because we know what it's gonna take to win our internal business and what it will take to win external business. And this is why Dave Zinsner, in that call, remarked that we're headed toward a goal of $8 billion-$10 billion of savings by exiting 2025, because we're so focused now on the right things, and that's just one example of the things that we will build out to make us a more credible foundry supplier.
Got it. That's, that's very helpful. And then I guess in terms of customer engagements, cases where you managed to land customers, what did they see in IFS?
To the extent there have been instances where things didn't play out the way you hoped them to, what were some of the reasons for the lack thereof?
Well, when we looked at the announced customer, we have MediaTek as an example, they saw in us a really hungry, aggressive partner, you know, that was willing to do what it took to win their business. Cost, quality, performance, you know, guarantees, collaboration. We put our best foot forward to go win MediaTek, and we learned a lot in that engagement process, similar to the Intel 3 customer that we, you know, that we won, but has asked at this point in time to remain unannounced. Where it hasn't worked out is people are just, they're waiting to see what happens, right?
They wanna believe that, you know, 18A is gonna be there, and so what we're doing right now is they're saying, "Okay, I want to see a few more test chips," or, "I want to see a more evolved PDK process design kit specification." I can tell you, all of them are interested in having Intel win. They're all interested in resilient supply chains. They're all interested in a U.S.-based supply chain. They're all interested in having choice, but they won't risk their business until we can meet the fundamental table stakes of the business, which is competitive power, performance, area, and cost. Once we have that, once we demonstrate to all of them, as we've demonstrated, for example, the one customer that, you know, we've got a prepay from, you know, last week, they gain confidence.
But there are others who are gonna be a little bit more risk-averse, so they want to see a little more silicon progress. One of our customers has a phrase that, you know, I often use called, you know, silicon speaks. In the foundry business, it's an incredibly measurable business, so having silicon, having them see the constant improvement in performance builds confidence such that they want to, you know, look at us as an alternate supplier to TSMC, which does a very good job for them. So, you know, we know we have to go do, we're knocking it out right now, and I you know, believe that we will win these customers, you know, as, you know, the months progress here.
Conflict of interest is something that comes up-
Yep
... in conversations with investors as it pertains to the foundry business. You know, TSMC, you know, pure foundry.
Pure foundry.
Nothing more, nothing less.
Yep.
Right? You guys obviously have a strong product group. Samsung has suffered from this issue as well.
How do you address that? You talked about separating manufacturing with the product group, but still under Intel, right?
Well, I think the way I look at it is that you imagine that I'm in front of Lisa Su or Cristiano Amon or Jensen Huang, or the people that you all know, right? How can I convince them that we're serious? The only way to convince them is you have to take concrete actions. For example, our sales forces are separated. We have two separate instances of SAP. We have firewalls, we have personal NDAs that our salespeople sign, that there are consequences if they violate those NDAs. We're super religious about protecting their IP. Pat runs two separate staff meetings. I do not attend the business unit staff meetings anymore, or do the manufacturing people. There are a few, few that we do combine, like overall HR issues, but we run the business separate.
We're creating two separate SAP instances to create this arm's length transaction, so nobody gets access to anybody else's data. And I think, you know, they have to watch our actions rather than our words, to see that this is happening. We know if we ever violate a customer's confidence, we're done in this business. Absolutely done. So we are religious about maintaining IP. By the way, we do a lot of work with the governments, you know, around the world, especially governments who don't like their names being used publicly, and we're gonna protect our information to those level of standards, right? That's how we have to think about it. And, you know, given the amount of energy we're putting into this and the investment we're making, into these areas, you know, I believe we'll pull it off.
But we all know that there are- there can be zero tolerance and zero mistakes in this, and we will do everything we can to firewall what we learn from our customers and make sure it stays separate from the IDM business. I think one way you could look at that, just as an existence proof, this morning, we announced that we're an investor in Arm. So if you want to know that we're serious about embracing that world, you know, which by the way, 80% of TSMC's wafers have an Arm processor in them. The fact that our organization, the IFS organization, is embracing Arm at this level, investing in Arm, doing partnerships with Arm, should give you a signpost that we're absolutely serious about playing in this business, because if you're not working with Arm, you can't be a foundries provider.
, just a couple of proof points as to how we do this.
. So on that point, you know, the partnerships that you've announced over the past, you know, call it 6-12 months, whether it be Cadence, whether it be Synopsys RISC-V , Arm, in terms of kind of building out that ecosystem, having table stakes to compete in the foundry universe, are you sort of at a steady state, if you will? Are you done kind of signing these partnerships, or is there more to do?
Oh, no, there's a lot more to do. I mean, there's a lot more to do with the existing partners. For example, you know, we announced something with Arm. We will do more with them, clearly, as they expand their base... you know, they have multiple interests in multiple areas, and they have been a superb partner. They're teaching us how to do, for example, low-power silicon, because they have 99% of the phone market. We're working closely with them in other areas of their business. You know, Cadence, we're working closely with. You know, we'll, we'll work to have some announcements from them, you know, over the course of the coming months. Synopsys, we just had this big announcement a few weeks ago, where they're filling out their IP portfolio, but it's also, how do we go to market with them?
How can we support each other? Because they view us as a new opportunity. There are a lot of third-party IP providers, folks like Alphawave, that we have an investment in today, right, who do some really incredible things with SerDes. So there's a lot more work to build out this ecosystem, and I think we're really just getting started. Our focus will be, for now, much more on Arm than around RISC-V, because that's where the volume's at, but expect more to come out in the coming months.
Okay, great. I've got a bunch more questions, but maybe I'll pause here and see if anyone has any questions in the audience.
This is not a shy crowd, usually.
I can keep going. Maybe on the long-term opportunity set for the foundry business, it's been a while since, and I think this may predate you, or you were part of the company when you hosted Investor Day, but you were not-
I was part of the company when the IBM PC came out, but okay.
Right, in 2022, I think the most recent Investor Day was hosted.
Pat and Dave shared a 2026 TAM estimate $160 billion for IFS, which implied a growth rate in the high single digits. Now that you're leading the business, does that still sound like the right context and format?
You know, it can, but I think the real issue is, you know, you have to look at it customer by customer. There are a lot of people who put out TAM projections, and, you know, the industry talks about $1 trillion of semiconductor business in 2030. I come from a sales background. I was in sales for, less than half my career, then till about 40% of it, where I just look at customers and can we win business? So we're very focused on who buys this stuff, what are the opportunities, and, you know, the forecast I have to build is more of a bottoms-up strategy. You know, we still have to get to this goal of how do we become the number two external foundry supplier by 2030.
But I start with, who are the customers and what do they buy, and where do we have the probability of winning? So you asked me a question earlier about how does the board evaluate me. The board is looking at the details, right? They don't want sweeping vistas of TAM. They want to know, who are your top five customers? How are you going to grow them? When can I see some announcements? And that's how they're measuring me and my organization, is are we bringing, you know, real meat to this conversation? So while I, I, you know, do, you know, care about long-term TAM forecasts, I think because of the size of the way these large customers buy, you know, one deal is really big.
Sure.
So you just have to really focus on the deals that are out there and closing them, and that's really what we're focused on, more than the TAM forecast.
Okay. And again, in terms of the opportunity set, you know, whether it be device type, whether it be end market, you know, is there a focus inside of IFS, or are you kind of looking pretty broad?
We want to look broad, but our heritage has always been one of high-performance transistors.
You know, you look at our business, our server business, our client business has always been more about performance and less about being power-constrained. You know, as a result, we have, I think, still a lot to learn in how to do great mobile designs. And some of the work we've done with Qualcomm as part of RAMP-C, some of the work we're doing with Arm, for example, right now, in terms of their mobile platform, we're learning more and more how to do a great mobile process. But our heritage is really high-performance silicon.
So you'll see us gravitate more toward more of the high-performance activities, and then also taking advantage of the packaging that supports, you know, that kind of technology characteristics, so very large die sizes, doing, you know, things like EMIB, which are ways, ways to connect dies on a single substrate, doing things beyond that going forward, offering our customers a five-year packaging roadmap. , that's where we, you know, I think, have a lot of natural fellow travelers, and then we're gonna build out our mobile process capability as we go along with the learning we're getting from, like, MediaTek right now.
So that's sort of a bias, if you will, to how we're thinking about it.
So you just talked about your packaging capabilities. Obviously, there's a shortage as we speak in advanced packagin i f you will, around CoWoS. Have you seen sort of an uptick in customer engagements as it pertains to your packaging technology? I think Amazon was one of your first customers-
Was the first one.
- to sign up.
First one. Amazon and Cisco both. We've seen a dramatic uptick, and I think one of the things, unfortunately, we don't have a camera here, otherwise, you guys could all see this on the screen, but I have, sitting in my hand, a single device on a single package. It's 100 billion transistors. It's 47 different tiles, 47 different active tiles, all on one common substrate on the back here. It mixes and matches Samsung HBM memory. It has TSMC-based I/O technology and compute technology. It has some of our compute technology on it, and this is the heart and soul of the Argonne supercomputer, you know, over in—at the Argonne National Laboratory.
You take 60,000 of these, another 20,000 Sapphires, and you have a machine that can do 2 exaflops a second of compute performance, so that's 2 billion billion floating-point operations a second. This installation goes live this fall. We believe it will be the fastest supercomputer on the planet, and the reason I bring this up in the context of packaging is this is sort of the acid test. If you can take 47 active, active tiles, mix and match this across different suppliers, have, you know, world-class back-end assembly test fields, so you're not throwing stuff out. This is why people are interested in talking to us about packaging technology, because we've done some of the hardest packaging things in the world and made them reality. So dramatic uptick.
I hope to soon announce what some of those customers are, but for now, they're going through all their evaluations. But we're doing a number of projects right now with a number of, you know, favorite names that you all know. We'll see how it turns out.
Got it. And then from a wafer processing standpoint, I had a question on how to think about your capacity expansion plans. We just discussed how you're building a wall-
Right
- between Intel product group and IFS.
Right.
I think some customers oftentimes want to see capacity first before they sign up, but you want to be prudent with your CapEx. So how should we think about your capacity plans from a foundry perspective? Like, what percentage of your total capacity today is purely dedicated to the foundry business?
, I mean, today, we're obviously super-
Right
... super small on a wafer fab capacity, and I'm not, you know, going to disclose by year what this looks like. What we are doing is, for any foundry deal, you have to commit capacity corridors. In fact, the phrase is corridor.
Once you commit that corridor, that corridor is inviolate to the customer it's giving to.
Okay.
So as we're working to strike deals, like, for example, like the prepay thing we just did, what's involved in that is how do we set up a capacity corridor for them or for any other customers? Once that's done, it's theirs, and the IDM business or server client business can't mess with it. Now, we have to be really judicious about is how do we forecast the size of the IDM business and the size of the foundry business? So we've created a fairly elaborate set of models where we look at what gets a corridor and what gets constrained to our internal business. And because we've been modeling PC business and server business for a long time, we have a pretty good handle on what that is, you know, plus, minus.
It's a fairly narrow band. The foundry deals will add capacity as we get customers in place, and part of our conversation with them is you know, getting sure we know the timing of that. We will then sequence our construction spending based on commitments from these customers and based on how well the client business and server business is doing. Take all that, model it together, that gives you the construction schedule for Arizona, it gives you the construction schedule for Ohio, it gives you the construction schedule for Germany, Ireland, and Israel, all five of those facilities.
And then the other way we think about capacity is the Tower announcement we just did this morning, where we have capacity in an older factory. We have tools for older technologies. We found a way to do contract manufacturing for them that take advantage of that extra space, those older tools that we weren't using, taking some investment from Tower to finish out the line, and now we're a contract manufacturer.
So by the way, being part of that business also teaches us more about how to be a better foundry supplier. So, you know, we never stopped talking to Tower, even after, you know, we didn't exactly get the approval we wanted, and we found a way to do business with them. I think that's gonna be very successful for both of us.
So, so this Tower announcement from this morning, so you're supporting them in 12-inch. Is that correct?
Right. Correct. Yep.
And then 8-inch-
They have their own facility.
They, they do, and you would produce overflow, or?
No, we've contracted with them for a very specific 12-inch flow for 65 nanometers that is in high demand by their end customers.
Got it. Okay.
Tower does some magical things with RF and power.
Right.
Power is super important, you know, for a whole bunch of applications. They are world-class in what they do. They wanted a U.S. factory. They wanted capacity expansion. We had to find a way to make it have economic sense for both of us, and because we have a building there, we have tools already there, we can bring more tools in as we need, and, you know, they add a few, you know, a few things of their own in, in this, you know, out pops a good 65-nanometer, you know, BCD process is how they coin it, for them to go use with their customers.
Got it. Okay. All right, that's helpful. We talked about the internal foundry model, extensively. I think starting in Q1 of next year
... you guys will report the manufacturing arm of the company, if you will a nd the product group arm, P&L, margins, et cetera. I guess my blunt question is: well, what is that gonna look like? The manufacturing arm you know, is subscale at this point, from an external-facing standpoint, but the P&L will be for both external facing and internal facing. Is that the right way to think about-
, think about it this way. We will sell wafers to the internal business units.
Right.
They will have a wafer price.
And we will recognize that price along with our costs in that P& L, and we'll also sell wafers externally. So the TMGF P&L-
Right
... will be a combination of both internal and external. You know, as to subscale, you know, we're driving, you know, still driving a lot of volume. It's not like it... It's a small operation-
Right
... but, you know, we will make sure those wafer prices that we're charging the business units are competitive with what they can see, you know, from an external supplier. So you'll see that all play out, and we're gonna take some aggressive goals, you know, in terms of margin targets. I'm not ready to disclose what all those look like today. I'll leave that to Dave Zinsner and Pat. But we have a pretty well-thought-out plan right now, and we're already starting to, you know, look at the business in terms of how they make decisions, both with an IFS format and with a classic Intel integrated format. We still run the company today on the classic integrated format.
But the IAO view is starting to inform how we think about things in the future.
Got it. And I guess in the, in the spirit of this true internal foundry model... your product teams, are they free to, like, leave Intel and go to TSM or, or Samsung? Or are they still encouraged to stay with Intel? Like, how is that?
gonna work? Like, how much
Well, I mean, it-
Discipline are you truly?
There are some rather, Grovian conversations that we're now having.
Okay.
You know, Andy was famous for having very blunt conversations, and those conversations are now happening. I think, you know, being part of an Intel business unit means you have all your capacity guaranteed, for all intents and purposes. So we're not gonna see a whole- you know, they won't be able to wholesale flock and give all their business to a foundry competitor, nor do they have to, because we're fairly competitive. But we do want them to have choice. And, you know, we have a very strong business with TSMC. We have a very complicated business with them, but a strong business partnership with them. And I am perfectly fine if our business units want to keep testing us to see how we're doing, because if we can't win their business, we can't win business externally.
Right.
So I think it's important for us to constantly market test internally and externally, make sure we're doing the best job possible, and yes, to some extent, the BUs will have freedom of choice. And it's incumbent on the factory team, you know, the manufacturing organizations, the R&D organizations, to do the most competitive process they can do to keep our internal business units whole. Because if we don't do that, our internal business units, the service and the client business, won't thrive.
Right.
We make the kind of margins that our investors are used to making, when we have process superiority. That's how this business has always worked. You know, if you look in the past, we were always returning 50%-60% gross margins, but we were continually a process generation ahead. We have the roadmaps in place to do that with Gate-All-Around, with backside metallization, and as we realize and ramp those processes, we believe we're gonna be able to give our business units more competitive advantage, and that will show up in their willingness to give the, you know, the manufacturing team more of their wafers. So that's how it works, but it'll be, it'll be good, healthy debate and conversation.
And I guess as you make this transition, do you lose any of the positives from an IDM business model perspective? You know, one team, one dream, design team, manufacturing team. I mean, that used to be the old sort of-
Statement, if you will, from Intel.
I think the dream itself is being recast into that. We all have positions to play on this field of play. Manufacturing guys have to have world-class cost.
You know, the business units have to have world-class architectures.
I think we're playing more effectively as a team by doing this. Now, we're having a few more arguments here and there, but it's bringing back healthy debate-
Okay, .
and choice and focus, and that's what we need to be to be successful. You know, it goes back to, you know, Pat saying, "The foundry business will make our internal business more successful and vice versa." In that continuous loop and knocking out, you know, five nodes in four years and, you know, working and getting a resilient supply chain, all those things benefit all of our customers, whether it be our business units or our foundry customers. So , I like, I like our chances, and I like the progress we're making, and hopefully, we're showing you enough in terms of signposts, whether it be, you know, the IP announcements we're making, the partnerships we're striking, the prepays that we're getting to show that there is progress being made and that, you know, our team hopes to keep bringing you more news like this going forward.
Got it. In the last, 60 seconds or so that we have, anything about the foundry business, anything about Intel broadly that, you wanna highlight or anything that we... I know, I know you don't spend too much time with investors, but anything that we collectively miss about the Intel story?
Well, I'm gonna be spending a lot more time with investors, as evidenced by today. No, I think the thing, and I'll speak of this, is, you know, part of the reason I came back, and this is, I'll end with a personal note. I think Intel is usually important to American manufacturing. I believe, you know, governments around the world view that. Certainly, the U.S. government views that. And I think it's critical that we remain successful, because losing Intel would be, I think, losing the crown jewels of American manufacturing.
And that's the thing that, if people ask what, what drives, you know, Pat, myself, the leadership team to want to come in day in and day out and drive this, it's preserving that capability for this country that I think is, you know, the mission that we've all taken on, and I believe the foundry business will go a long way in helping us get there.
Great. It's a great place to end.
Thank you so much.
Thanks, everyone.
Appreciate the time.