Okay, good morning, and again, welcome to the first day of JPMorgan's 53rd Annual Technology, Media, and Communications Conference. My name is Harlan Sur. I'm the Semiconductor and Semiconductor Capital Equipment Analyst for the firm. Very pleased to have Dave Zinsner, Executive Vice President and Chief Financial Officer of Intel, here with us today. Dave, thanks for joining us this morning.
Thank you for having me.
Yeah, I think a good place for us to start off with is with Lip-Bu, Lip-Bu Tan. Intel's relatively new CEO has been in place now for about 60 days. Prior to that, he was a board member at Intel for two years, obviously widely respected in the industry as a successful venture investor, spearheaded the turnaround at Cadence Design as CEO for 12 years. Dave, you know, as interim CEO prior to Lip-Bu joining and CFO for the past three and a half years, you've obviously been working very closely with Lip-Bu. From an outsider looking in, though, it does appear that Lip-Bu is not making any massive changes to the strategic direction, product, or manufacturing roadmaps that were in place before he joined. Right?
He acknowledges Intel's historical profile as a technology and compute powerhouse and wants to, from my perspective, number one, return Intel to that leadership position in core compute with best-in-class client and server products. Number two, expand the leadership in compute to take advantage of the rise in AI compute workloads. Three, from a manufacturing perspective, return Intel to a leadership position in advanced leading-edge manufacturing, but then leverage that, right, to grow a world-class foundry business. Is that a fair assessment of the strategy so far? Through his first 60 days, I know you've been working closely with him. What are the key observations, strengths, and weaknesses of the organization?
Okay. All right. Just since, you know, I might be making forward-looking statements, I ought to say, you know, just a safe harbor that we are, you know, if I do make forward-looking statements, you should refer to all the risk factors in the SEC filings as filed. Yeah, look, I think it's a fair assessment that Lip-Bu isn't thinking about massive changes. I think, you know, when he looks at the business, what he feels is the biggest issue at this point is a lack of execution. Yeah. You know, he is attacking that in a number of different ways. He is really flattening the organization. You know, we're going to go through a process to flatten the organization in aggregate for Intel, but he's already done that at the executive team level. I mean, he's, you know, I don't know, like 15, 17 direct reports.
He's got a lot more engineering leaders reporting directly to him as opposed to, you know, having an intermediary report between them. I think this, like, flattening the organization is important to him because what he wants is the lowest level in the organization to be closer to him and so that he's hearing the good, the bad, the ugly of what's going on so that he can help address those. I think through that, I think we'll see a lot of improvement in terms of the execution. He's also spending a lot of time with customers. Lip-Bu's Rolodex is like nobody else's, I think, in the semiconductor industry.
You know, I think what he feels was lacking in addition to execution was not enough listening to the customer in terms of what they really need and then going and developing products to meet their demand, their requirements on a long-term basis. He is all in on that. I think he was telling me he, like, his longest day was 22 meetings, I think he said he had. I think it was on a weekend. He is working really hard to go address all of that. Maybe perhaps the one area where I think he will spend more time that could potentially evolve to a different strategy or at least a pivot on the strategy is in AI.
You know, I think as I look at the leadership of the organization, you know, we all have our roles, but I think as it relates to AI, I think of Lip-Bu as actually the leader of our AI initiative. And he spent a lot of time investing in that space, you know, through his venture fund. And so he understands it at a level that I think is unique for a CEO in the semiconductor space. And so he is really rolling up his sleeves to make sure we're developing products that uniquely meet what isn't getting met today from an AI perspective. And, you know, obviously power is a big issue right now. The cost is a big issue. You know, as you move away from training into inference, as you move away from cloud to on-prem to edge, there are different requirements.
What he wants to do is find those opportunities with the customers where we can uniquely develop solutions that meet what they need. That's probably the only area where I see a meaningful change in how he thinks about that part of the business. Outside of that, I think you're right. I think the strategy is the strategy. It's we got to execute better.
Yeah. I think, as you mentioned, you know, and he mentioned this on the call last month as well, one of the biggest takeaways, learnings, was a need to fundamentally transform the culture and the way in which the organization runs, eliminate excess bureaucracy, the need for a flatter organization, smaller teams, stronger decision-making authority, very strong focus on execution, right? Which implies accountability and transparency. As you mentioned, phase one of the strategy resulted in lower OpEx targets for this year, $17 billion, $16 billion next year. Also cut the net CapEx outlay by $2 billion- $18 billion. On the organizational side, how is he implementing a more nimble execution-focused mindset across the organization, right? What division teams are reporting to Lip-Bu? What has he changed in terms of KPIs, targets, incentives? Because from my perspective, historically, like, this is where prior Intel CEOs have faced challenges, right?
Yeah. Yeah. Part of it is what I already said. He's flattened his organization. He now has the two major businesses, data center and client, reporting to him. He has three different engineering leaders on the product side reporting to him. He has sales reporting to him. He has operations separated from sales now reporting to him. He'll have government affairs actually reporting directly to him. We have a leader over top of manufacturing and TD that, you know, runs, let's call it 90% of the foundry business. That person reports directly to him and then the host of the G&A functions. That's different in terms of the structure. It's way closer to decision-making. He wants to, you know, he wants to be able to understand at a very detailed level exactly what's going on in any part of the organization.
That's kind of, I think, the first cultural step in the culture change that he's implementing. The second, which I also said, he's collapsing the number of layers in the organization. That, you know, has that benefit that I talked about, which is greater transparency, but it also forces kind of less overhead, less bureaucracy in the decision-making, which is part of what has plagued us is a lack of quick decision-making. When you have less layers, you have less people that have to check, you know, and sign off on something. Things move more quickly. You know, we're more nimble in terms of our approach.
You know, the cost, while the cost reduction, I guess, was in some ways important in terms of improving the P&L, it had more to do with the cultural change of like, hey, we're going to reduce the amount of people making decisions. We're going to reduce all the, you know, the check-the-box activities that happen in the company. We're going to get rid of all of those things. We're going to operate more like a really big startup as opposed to, you know, this big Goliath of an organization that, you know, just crumbles under its own weight. He has gone through a lot of that. It'll take this quarter to kind of work through, you know, what functions need to streamline, what areas are not value-added.
Another thing that he did, which, you know, I know a lot of companies have done, so it may not seem that transformational, but actually is pretty important for Intel, in my view, is he has this return to office. You know, we did have a number of the team that was not working on site full-time. I think to some extent that has inhibited us and impacted our ability to execute on, you know, roadmaps and process. You know, just forcing that mindset that we're all going to be back in the office rolling up our sleeves to get things done, I think was an important message to the organization. What I also think has hurt us is, and this has been kind of like the Intel model, is we've somewhat developed the technology independent of what we're hearing from customers.
You know, with kind of an assumption around what we think is needed. We heard a lot from customers that we were bringing out products that were not addressing all of their needs. You know, probably we were not taking enough of the information that customers were telling us back to inform us in terms of how to, you know, what we should do from a development perspective. He has leaned in a lot. It is partly him going, you know, he is going to see customers, he is going to see CCEOs and, you know, the major systems engineers in the company to understand what their requirements are. That, of course, is something he is doing and that is important. It actually goes beyond that. You know, we have generally been, you know, kind of a three-layered to customers.
We have a sales force, then we have a, like, a product management team, and then we have the engineers behind them. It was not the default to send engineers to customers. I've seen that in my past and other companies I've been CFO of, that when you allow the people designing to go talk to the customers, they hear information in a way that, you know, the sales team, for all their strengths, just don't pick up. You know, and it's like, you know, actually customer told me that we need A, B, and C, but based on the way they were talking about their system and the roadmap, they actually don't know they need EEFG. I'm going to develop EFG into the product because that's going to actually meet their requirements in a way they weren't even expecting.
Moving the engineers closer to customers, I think, is a big cultural move for us. Now, as it relates to KPIs, which you asked about, you know, obviously, you know, what we'll be looking at is can we develop products in, you know, first silicon, you know, A- steppings. We have not done that in the past. You know, is the process, you know, yielding in a way that, you know, follows along a curve that we expect. Is performance of the wafers following along the curve that we expect. We'll be measuring all of those things clearly. I think most importantly, we're going to measure that we're actually meeting what customers are asking for and they're excited about the products that are coming out.
You know, normally I would start off the discussion on your product portfolio, but I want to start off with manufacturing because the challenge here for Intel is twofold. One is drive back to leadership position in leading-edge process technologies, right? Because leading-edge process technologies drive a big part of the performance of the core compute portfolio, right? Number two is that you can then take your leading-edge technology to build and then leverage that to build a world-class foundry business, right? Again, looks like no change to process technology roadmaps that were laid out by Lip-Bu. But, you know, what is he doing differently to accomplish this twofold strategy, right? Maybe use 18A process technology as an example. What is the team doing to drive differentiation, better economics on your core products like your upcoming PC client CPU product called Panther Lake? What are you doing to attract more external foundry customers to adopt 18A?
That's a wholesome question. Yeah. Look, you know, in a lot of ways, this is blocking and tackling, you know, the foundry business. And, you know, while it's felt, you know, relatively challenging as we progressed to get caught up on process, we've actually made a lot of headway. You know, unfortunately, to get from where we were to get to a point where we have 18A ramping in the fab is a, you know, painful process. It requires a lot of capital investment. It's never a straight line. I've been through this before in my prior company. You know, catching up on process is a really hard undertaking. While it's not to where we were driving, when you look back, it's actually a fair amount of progress that we have made.
When you look at 18A, it has differentiation in it, you know, which, you know, to your question, how are we driving differentiation? You know, having, you know, Gate All-Around, what we call Ribbon FET together with backside power, which we call PowerVia. You know, that's actually a pretty innovative solution. When I, you know, for the split second, I was acting in CEO and they actually let me talk to customers, I heard a lot of positive aspects about that, about backside power with Gate All-Around and how integral it's going to be in the future, particularly in high-performance compute. You know, one customer told me, "We will never go back from backside power." You know, the fact that that technology is available to us is critical to our roadmap. I think, you know, we are making the progress we need to make on process.
Now, you know, the one thing about 18A was it was developed initially as just something for Intel. And we intercepted it relatively early and it allowed us to develop PDKs for the industry. But it still was not from the grounds up developed as a foundry node. 14A is developed from the grounds up as a foundry node. It will have PDKs that are exactly comparable to what the industry would expect from us. You know, a 1.0 PDK for 14A is going to look exactly like a 1.0 PDK to our competitors at that given process. I think we are at a point where we have driven a lot of maturity.
You know, and, you know, Lip-Bu, what he will spend a lot of time on and maybe something that has always been a challenge at Intel is as information filters up to the CEO, it gets filtered a lot. Yeah. And tends to have been presented in the most positive light. Even when Lip-Bu was on the board, he was really good at, you know, kind of filtering out some of that spin and really understanding at the detailed level exactly where we stood, where we stood on yields, where we stood on performance, you know, how we felt like the wafers would ramp, when we would get to mature yields and so forth. He spent a lot of time understanding that even at the board level, at the CEO level, it's like to the power of 10, basically what he's doing.
I think there we will all have a very clear understanding of where we stand. And when you have a clear understanding of where you stand, you have a clear understanding of what you need to do to improve, you know, improve the process. I think that, to me, is the major focus of Lip-Bu's as we get into the Intel Foundry.
Yeah. It was good to see, it was good to see the good participation at your recent direct foundry event, I think held in Santa Clara. It was also good to see, to your point, right, that Intel has always had a strategy of we develop our processes for our products, right? Then we take this process and whoever wants to use it as an external foundry customer, go ahead and use it, right? It's never been with the view of what do our customers really want, right? What really want, right? Now it's, you can see the strategy shift, right? You're developing, like you said, for 18A and for 14A. You've got several different flavors, right? Low power, high performance, and so on. The same type of offerings as your world-class, you know, foundry competitors.
The services aspect of foundry is moving more and more towards like, well, what your customers would expect. That being said, you know, on 18A, you've had some traction, right? And previously, I think the team articulated that you were on track to tape out your first external design on 18A in the first half of the year. Did the team execute to that? And then whether it's 18A or 14A, I mean, how many customers, external customers, have committed to Intel Foundry? And, you know, are they a mix of fabless semi-customers, cloud and hyperscalers, OEMs? Like any color there would be great.
Yeah. The first 18A customer is going to be Intel products. It's Panther Lake, and the first SKU is expected to be out by the end of the year. That is our first win, so to speak, if you put on the Intel Foundry hat with 18A. And Really, we always expected that, you know, the predominant amount of volume of 18A was going to be internal to us. You know, we would win customers here or there that would fill in the gaps. We do not need a lot to fill in the gaps to make 18A a good node from an NPV perspective. I feel actually pretty good about our ability to drive a reasonable return on 18A. 14A, you know, obviously it gets more expensive. At present, it's expected to have high NA and, you know, that's a more expensive tool.
You know, I think we do need to see more external volume come from 14A versus 18A. You know, so far, you know, and we've talked about it in the past. We have, like, you know, the traditional, like, pipeline modeling, a bunch of potential customers, and then, you know, we get test chips, and then some customers fall out in the test chips, and then there's a certain amount of customers that kind of hang in there. Committed volume is not significant right now, for sure. You know, I think we've got to partly prove ourselves a little bit with our own product and eat our own dog food here. I think it's more likely to come that we, you know, we start to see some engagement around customers. The conversations have been good.
I would tell you that every customer, surprisingly, you know, at one point, you know, under COVID, everybody was worried about the supply chain. You know, everybody kind of forgot about COVID and the supply chain concerns dissipated, I would say. I'd say it's now back to the more and more I hear customers wanting to have a second source. It doesn't mean that the competitors, or the largest competitor, won't get, you know, every bit of share that they want. I think there's always a requirement for a second source in these cases. That's the opportunity and that's the area we want to fill. We also have, I think, really good packaging, and we tend not to talk about that as we think about foundry, but we'll have revenue in the back half of the year on packaging, advanced packaging.
I think that's a great opportunity for us. The technology's differentiated already. It's mature in a way that, you know, we still have work to do on the wafer side. It's a good vehicle to get customers in the door working with Intel Foundry in an area they have a lot of trust with us and then kind of work them into foundry deals over time. I'd say the other thing is, you know, the challenge of being a foundry player in our construct is that the products business competes against, you know, many of the customers that could be potentially customers of foundry. What we have to do is make sure they feel confident that, you know, IP is getting protected, but maybe more importantly that supply is protected, you know, that we're not going to disadvantage them relative to products in terms of supply.
I think Lip-Bu in that regard is a real asset. He's trusted, you know, he does the right thing. If he commits to a customer, he delivers on that commitment and c ustomers trust that he will do that and t hat, I think, has changed the dynamic of the conversations we've had with potential foundry customers, just literally based on him walking in the door and making them feel at ease about those concerns.
From a financial perspective, I believe the target is still to drive Intel Foundry to operating profitability breakeven exiting calendar 2027. Remind us on the parameters needed to achieve that, right? Mix of internal versus external customers, process technology mix. And by calendar 2027, like, what do you anticipate your external wafer usage as a percent of total wafer shipments will be? I think today it's about 30%.
Yeah. Okay. So we still feel on track to hit breakeven sometime in 2027. You know, I think when we committed to it in 2024, we said, "Hey, it's going to be somewhere between 2024 and 2023." Most people kind of settled in that that must mean 2027. And that's generally kind of what we're thinking is we can be breakeven. It doesn't require a ton of revenue for foundry. It's somewhere in the, you know, single digits, called low to mid single digit billions of revenue that foundry's got to get from external sources. I would just remind you that some of that's going to be our partnership with UMC. Some of that's going to be our partnership with TOWER. Some of that's going to be packaging. And some of that's going to be 18A.
Some of that actually is going to be older generations, you know, like for example, Intel 16. It is not a ton that has to come from 18A for it to work. What is important is that we drive the right amount of volume from Intel products through foundry. We are fairly optimistic that we will be able to do that. I feel highly confident that we will breakeven in 2027 on the foundry side. What was the other question, Harlan?
The other one was about Intel on your core products, external versus internal mix. Yeah. Today, I think it's 30%.
It's peaky, but I would tell you that we will have a fair amount of volume at external, you know, mainly TSMC on a go-forward basis. You know, products has some latitude as to how they design products. We go through all of that as a team. Of course, we want to see a fair amount of the volume come from Intel Foundry. We also recognize there are unique situations and performance requirements that would drive you to another source of the wafer. We want that. We want that because we want to drive the best products ultimately for our customers. We also feel like it's a healthy dynamic to have foundry feel like it's competing for every wafer with Intel products. You know, I think that drives better performance out of Intel Foundry as well.
I'd say lastly, you know, when I was at Analyst Day and what was that, 2022? 2022 when I came in, we had this concept. We do not talk about it a ton now, but we had this smart capital concept. I remember Harlan, and it had like all these things that we would use to kind of, you know, enable us to, you know, stand up a foundry business, you know, given the capital requirements to do that. One of them was we would have a balance of internal and external wafers for Intel products. We would leverage the foundry network beyond Intel Foundry to provide supply and t hat helps offset some of the capital spend and keeps us in a good place from a cash flow perspective. It is somewhat aligned with our capital strategy on a longer-term basis.
Is there a right way to think about the mix, maybe 20-30%?
Yeah, something like that.
Okay. Let's turn over to the core products. You know, in core, internal CPU products and internal product development efforts, given Lip-Bu's familiarity and expertise on chip design implementation, you know, has he assessed Intel's chip design and verification flows, IP development capabilities, software development execution, right? Where does he see the opportunities to improve the quality, efficiency, execution of Intel's chip design infrastructure?
Yeah. I think what Lip-Bu recognizes coming in is the strong x86 franchise that we have. I mean, in a lot of respects, you could argue that our share should be a lot lower than it is. It's not because we have a really strong ecosystem, and the moats around this business are quite significant. You know, I was talking to one of my banker buddies, and he was telling me that that bank doesn't even think about moving away from Intel, given the network and the ecosystem and the support that they get. I think that is our advantage that we have that. Quite honestly, you know, it's a little bit embarrassing that we've developed products that haven't met customer requirements to the level that we should because it would be really easy. We could roll out of bed and win market share.
What Lip-Bu now is trying to make sure we do, which I somewhat talked about before, is make sure that we are developing products that meet customer requirements. Because if we have a strong ecosystem where customers want to be engaged with us and we are developing products that are meeting their requirements, we are going to do quite well over time. That is primarily what he is focused on, making sure we are taking customer input to develop products. I would say beyond that, one of the things that he has identified, which we have known about, is that we do not put products out on an A-stepping, you know, on the kind of first attempt at a product.
In fact, I think Sapphire Rapids came out on an E-stepping, which is, that's a significant amount of time, you know, going back and forth between design and production to bring out wafers to figure out, to get a wafer that actually worked for customers. He wants to shift left significantly. He basically has mandated we are not bringing out products that aren't ready, that haven't been simulated and tested. Because we spend, partly it's just quality, partly it's you spend a whole bunch of time then fixing it afterwards on the back end when the products aren't out, you know, first time with success. He is shifting left significantly in terms of the development. I think you will see, and maybe back to your KPI question on what should we be looking for, I think one thing you should look for is that he's attracting talent.
You know, it's well written, and I think there is some truth to the fact that we have lost talent at the company, and we do need to rebuild that. The great thing about Lip-Bu is he's a magnet for talent. I think you will start to see people come into the organization at the, you know, senior engineering levels that are, you know, world-class that will help us, you know, in terms of improving our design process and ultimately the architecture of the products. That to me, I think, is the biggest thing you'll see in terms of change.
In Client PC, you know, the big intro this year is the ramp of your lead 18A product for PC client called Panther Lake. It was clear that the team slightly downticked its last earnings on the ramp of Panther Lake from what used to be second half of this year to at least one SKU exiting this year. What happened? Is this more of a product delay, in other words, design-related issue side, or is this more that the 18A process technology and manufacturing node will be quite ready?
Yeah. I think it's a combination. First of all, the market, you know, usually in the way you work this is, and what we've done in prior generations is we've brought out a first product, you know, towards the end of the year in advance for CES and You see a bunch of other products ramping through the first half of the following year. That's kind of the way we did Meteor Lake, That's the way we did Lunar Lake. So Panther Lake's kind of somewhat following along our typical process in that regard. Now, probably if we could have driven harder to get products out more quickly, we would have, you know, we had a drive-through that was a bit more aggressive, which is somewhat why we left it open to back half of the year. It could be anytime back half of the year.
We knew we'd at least get some products out or a product out by the end of the year. We're, you know, in some ways hoping that we could bring some things out earlier. You know, it just didn't happen in the way we were expecting. That said, you know, we feel pretty confident that 18A will, you know, be out with a product this year. I feel very confident that, you know, as we progress through next year, we'll see that ramp in the mix, you know, drive towards more Panther Lake. That will be helpful next year from a margin perspective, by the way, because we get a twofer for that. One, we, you know, move wafers that, you know, were mostly done externally. We move some of that back internally. You know, obviously Panther Lake has 18A. That's internal.
That helps us a lot in terms of the profitability over time. Maybe even more importantly, Lunar Lake, because of the way it was designed, has memory in the packaging, and we essentially pass that through. We buy it at the same price we're charging the customer. That's not the architecture in Panther Lake so As we, you know, move into Panther Lake, the margins just pick up, you know, from that perspective too.
On server CPU product and platforms, launch of Granite Rapids last year was a very good step in the right direction, right? You're shrinking the competitive gap between you and your competitor, but there still is a gap, right? Next in the lineup is Diamond Rapids. What's the timing on Diamond Rapids? What's the early feedback from customers? When does Intel really fully expect to sort of close that server performance gap with your competitor?
Yeah. We have not talked about a date for Diamond Rapids, so I'll hold off on that. The product marketing people like to wheel that out at a particular moment in time, and they get mad at me if I jump their gun here. You know, early feedback's been good. You know, it does close the gap more in terms of the performance. I would just say, you know, our ability to maintain the share we have gets back to that same comment around ecosystem. I mean, the fact that we have maintained a significant amount of share has been a function of the fact that we do have great, you know, great support, great ecosystem, great stack, and so forth in the data center space. You know, we still have work to do, I would say, on data center.
You know, at the end of the day, we've got to develop something that gives customers the best perf per watt really and That's the measure until You know, every iteration of the product has gotten us closer. You know, there's a Farallon product after Diamond Rapids that does an even better job at driving better performance.
Last year, on the financials, last year you guys talked about free cash flow breakeven this year. Then you kind of pushed that out to free cash flow breakeven exiting this year. Last earnings, you talked about focusing more on operating cash flow. How should we think about your operating and/or free cash flow potential this year? How do we think about it over the next few years?
Yeah. We have, you know, backed off a little bit on the cash flow for this year, mainly, you know, given the uncertainty with tariffs and so forth. It's just a little unclear as to how things will play out for the year. We are very focused on it, I would just say. You know, areas that are in our direct control, like capital spending, like operating expenses, like working capital, we are driving to improve the cash flow. We also are going to augment our cash flow this year with the sale of 51% of Altera. That will help out and allow us to delever this year on the balance sheet.
Beyond that, you know, my expectation is, as we, like I said, as we move into Panther Lake, you know, there's a better cost structure for Panther Lake from the standpoint of margins as we drive, you know, better execution on the product portfolio and see that translate into growth. As we move more and more to the advanced nodes, which have, you know, a cost structure that isn't that different from the cost structure of the older nodes, but yet carries a much higher ASP on a per wafer basis, we see a meaningful improvement there. We get the OpEx down. We drive CapEx efficiencies. You know, our cash flow should meaningfully improve.
Perfect. Appreciate your participation, Dave, today, and look forward to monitoring the progress of the team this year.
Thank you, Harlan.
Yeah. Take care.
Thank you very much.
Thank you.