Intel Corporation (NASDAQ:INTC) New Segment Reporting Webinar Conference April 2, 2024 4:30 PM ET
Company Participants
John Pitzer – Corporate Vice President of IR
Pat Gelsinger – CEO
David Zinsner – CFO
Conference Call Participants
Timothy Arcuri – UBS
Joe Moore – Morgan Stanley
Vivek Arya – Bank of America Securities
Aaron Rakers – Wells Fargo
Vijay Rakesh – Mizuho
Srini Pajjuri – Raymond James
C.J. Muse – Cantor Fitzgerald
Operator
Thank you for standing by, and welcome to the Intel Corporation New Segment Reporting Webinar. At this time, all participants are in a listen-only mode. After the speakers’ presentation, there will be a question-and-answer session. [Operator Instructions] As a reminder, today’s program is being recorded.
And now, I’d like to introduce your host for today’s program, Mr. John Pitzer, Corporate Vice President, Investor Relations.
John Pitzer
Thank you, Jonathan, and welcome everyone.
At the close of market today, we published an 8-K containing full year recast financials for 2021, 2022, and 2023. The recast is the culmination of over a year and a half of internal work to create the systems and processes needed to establish a foundry like relationship between Intel Products and Intel Foundry and is aligned to the operating segments we officially adopted at the start of the 2024 fiscal year.
Today’s webinar will unpack the recast financials to provide a better understanding of our vision for Intel Foundry, review in-depth the underlying economics of Intel Products, and explain how each will help drive improved operations and financials going forward.
I’m joined today by our CEO, Pat Gelsinger, who will outline our vision for Intel Foundry; and our CFO, David Zinsner, who will discuss Intel Products and why we are confident our new reporting structure will be a catalyst to overall profit improvement and value creation over time.
After our prepared comments, we will have time to take your questions. We are in our quiet period and as such, we would ask that your questions today relate to the materials covered in the webinar and our recast financials. We will report our Q1 earnings after the close of market on April 25, and we’ll be more than happy to address near-term questions about the business at that time.
Before we begin, please note that today’s discussion does contain forward-looking statements based on the environment as we currently see it, and as such are subject to various risks and uncertainties. It also contains references to non-GAAP financial measures that we believe provide useful information to our investors.
Our most recent earnings release and annual report on Form 10-K and other filings with the SEC provide more information on specific risk factors that could cause actual results to differ materially from our expectations. They also provide additional information on our non-GAAP financial measures, including reconciliations where appropriate to our corresponding GAAP financial measures.
With that, let me turn things over to Pat.
Pat Gelsinger
Thank you, John, and welcome to today’s webinar.
Three years ago, we began this ambitious transformation that we called IDM 2.0.
First, rebuilding the execution engine of Intel. Second, reestablishing process and product leadership. And finally, becoming a world class foundry and standing up our foundry services. And I’m proud of what we’ve accomplished today.
We’re closing in on the finish line of 5 nodes in 4 years. 18A is manufacturing ready at the end of the year, and the first Intel designs are in fab and we already have 5 external customers committed to 18A. We’re already well underway on the next generation with 14A and beyond. And we have a building momentum in our AI leadership products with AI enhanced Xeons.
We launched the AI PC category. We’re strengthening our position in accelerators and our AI everywhere strategy, the PC, the edge, the on-prem data center in the cloud gaining momentum. And we’re the only industries, systems, foundry for the AI era. That’s resilient, trusted, sustainable, and a global manufacturing footprint. And today, over $15 billion of lifetime deal value is already committed to that foundry strategy.
Today, an important next step on our journey where we’re separating the financial reporting of Intel Foundry and Intel Products, and we’re providing a P&L view to our manufacturing engine for the first time ever in our history.
And we believe this transparency and accountability is needed to drive best-in-class cost structure across both groups to complement this return to technological leadership. Our goal, the exceptional stewards of our owners’ capital while creating value for all of our stakeholders, our customers, our employees, our shareholder, and importantly, all the communities in which we operate.
There are a few key points, we’d like you to walk away from our discussion today. The reasons behind our strategic shifts from IDM 1.0 to 2.0. We see this clear path to driving Intel Foundry to break even over the next few years. And what we see as a self-sustaining model with solid returns for Intel Foundry and the consolidated company by 2030.
IDM 1.0 was an incredible strategy while it worked, and it enabled us to deliver the world’s most important technology and deliver superior returns for many decades. And it was fundamentally rooted in our ability to maintain a 2 plus year process technology lead. We invested to out invent everyone through proprietary IP and process technologies.
We were optimized for speed of node transitions and always being able to shift just one more wafer. And the substantial majority of the free cash flow that the tools generated in our fabs were in the first few years of their service. But Copy Exactly! left no room for cost downs or long life assets to occur over time, and we chose to monetize our intellectual property, both products and process singularly through the products that we delivered. At our best, Intel Foundry cash flow from operations only equaled the CapEx that we put in place.
But the middle of the last decade and beginning with the need for EUV, there was an incredible growth in the capital intensity and this necessitated extending the life of both nodes and the tools that delivered them.
And even though we helped invent it, we were slow to adopt EUV. We were slow to adapt to the step-up in capital intensity and this fundamental shift to a foundry model. The U.S. and the Western world became far behind our Asian peers and this is what has driven us to respond with our IDM 2.0 strategy.
IDM 2.0, we will continue to drive Moore’s Law of Advances on both the wafer with gate all around and backside power with lithography with High-NA, but also critically in advanced packaging. And now Moore’s Law goes from X and Y to the Z dimension or truly innovating with Moore’s Law in all three dimension.
We are also making fundamental moves to standard intellectual property blocks and processes, where we became so disconnected from the industry, we were able to leverage the industry. We are also addressing the rising capital requirements by optimizing for cost efficiency and extending the life of our assets. We are addressing the scale challenges by bringing in external designs into our factory network and uniquely able to capture margin stacking and monetizing the IP in both our products and our foundry.
Today, is an important incremental step. This journey began three years ago when I returned and we’ve made great progress and we highlighted this at the Direct Connect Event, sort of our coming out party for the foundry industry. Five nodes in four years. Intel 18A on-track and manufacturing ready at the end of this year. And our first products in manufacturing with Clearwater Forest and Panther Lake already in fab today.
But we’re not stopping there. We also, at Direct Connect, introduced 14A, the next process node after that. We also expanded this node build out, bringing process variations to fill this picture beyond five nodes in four years. A full portfolio of leading edge, but also filling in more mature nodes, extending our capital assets with our partnerships with UMC and Tower Semiconductor.
We are also, after this intense five nodes, four year period, returning to a more normal two year cadence for leading edge technologies and providing a meaningful benefit to startup cost and efficiencies.
This is an impressive picture, where as we have reengaged with standard practices, the IP ecosystem and most importantly, the EDA customers, we’ve quickly built a vibrant ecosystem. Nearly 300 companies were at Direct Connect, including almost all potential fabless customers showing up there to see what is now possible.
And as you see in this picture, 18A covered top to bottom and will be shortly releasing the 1.0 PDK. And already impressively, 14A is already well underway to a complete solution set. Light years ahead of where we were on Intel 3 or 18A, we are already well advanced in how the industry is embracing this road map of our foundry offerings.
You’ve seen our say/do ratio has been very high. We’re addressing each shortcoming of IDM 1.0 in the EUV age. We’ve demonstrated clear and consistent progress. I can confidently say the required transformation is now well underway. The foundation of transformation has been the improving profitability and this regaining of transistor performance and cost.
This is a picture of Intel’s benchmarking versus the best of foundry alternatives. As you see with Intel 7, we went through great steps to avoid the need for EUV. And as you see, it produced deficient in power performance area and cost. Intel 3, a major step forward in power and performance. And we improved significantly our area and EDA and cost, but still gaps to the best in industry.
So Intel 3, a big step forward, but more to do. 18A is there, a bit ahead of competitors in time and delivering on power, performance, area, and major improvements in cost. 14A, the first embrace of high and a, and we expect to be ahead on time and on all metrics, power, performance, area, and cost across an entire spectrum.
And as you see in this picture, we’re also, for the first time, optimizing for both mobile power and for high performance computing out of the gate, truly making a comprehensive offering for the foundry ecosystem. But this last row was super powerful.
In packaging, we’ve gone from strength to strength. And with that, this idea of chiplets allows us to even make greater use of Moore’s Law. We’re able to, earlier in their life, have smaller die, which makes the process technology available for server and AI chip designs, also using less of the dye for I/O and analogs for more efficient use of leading edge technology overall.
And this to me is one of the key aspects of keeping Moore’s Law alive and well. And speaking of Moore’s Law, as we look at this picture, you know, the magic of Moore’s Law has always been giving power, performance, area, and cost.
As we went through our stumble of not embracing EUV and Intel 7, this allowed the industry to catch up. But the cost and complexities of EUV resulted in a flattening of the curve for the entire industry. We have now broken through that EUV wall and turned the corner toward High-NA. And importantly, we’re putting the economics back into Moore’s Law, where it’s going to be uncompetitive for earlier pre EUV technologies and for chips and fabless vendors using those to not embrace the leadership technologies.
We’ve seen through the industry’s history that once a decade, capital technical complexity reduces the number of firms capable of economically manufacturing leading edge transistors. EUV adoption is this decade’s barrier, and AI is driving an explosion in compute needs. Physics is driving density and package and bringing more of these chiplets closer together and to package them in 3D. The rack is becoming a system. The system is becoming a chip, and as we get to the other side of this EUV transition, as you see in this picture, hugely valuable position for Intel in the industry.
Intel Foundry is going to drive considerable earnings growth for Intel over time. 2024 is the trough for foundry operating losses. We’ve committed to being the number 2 foundry by the end of the decade. And between now and then, we’ll hit breakeven operating margin about midway through that and then driving operating margin improvement through the period.
Leadership technology, stabilizing investment, cost and capital efficiencies, these are the tools to accomplish that financial outlook.
First and foremost, Intel Foundry pricing and margins reflect this process technology deficit that we see today. The shift to EUV nodes moves from deficit to parity to leadership, and this results in pricing margins that leadership commands. Until 18A and beyond return to structural cost parity, and this is already built into the product P&L’s that we’re modeling for the rest of the decade.
But this also presents an important factor for our external foundry customers as they consider Intel Foundry. We’ve already derisked the technology, the ramp, the defect curves by the volumes, the significant volumes of Intel Product wafers going in. And as we finish this period of 5 nodes in 4 years, we stabilize our investment portfolio.
We also see that Intel products made a pragmatic decision to utilize this external foundry capacity over this period of time where we stumbled. But we are more incentivized than ever to bring these wafers home. The road map decisions are made. The plans are in place.
And over this period, we’ll be approaching 2 full fab modules of capacity that we’ll be bringing home. Our smart capital strategy is well underway. This is a key piece of our profitability journey, and we’re delivering on all aspects that we’ve laid out.
Recently, our CHIPS announcement, absolutely huge, $8.5 billion of grants, $11 billion in loans and greater than $25 billion in tax incentives. Our EU grants, well over $10 billion committed, Israel, Ireland, Germany, Poland and others, great progress on having those in place. We established our First SCIP at almost $15 billion, this ability for others to partner in our capital investments. Literally, we delivered on all aspects of this deliberate strategy.
And in the process, we have set in motion the rebalancing of the global supply chain. The world will benefit from this work for decades to come. Capital offsets provide a powerful tool to address the geo gaps but also improve capital efficiency. As we move past five-nodes-in-four-years as our EUV fleet is operational, we expect a more normal capital cadence. And Dave will cover further end-to-end efficiencies that we see resulting from the Intel Products, Intel Foundry separation and transparency.
We see that we’ll be approaching double-digit ROIC by 2030. All 3 drivers largely, if not entirely, are within our control, and the journey doesn’t end there. External foundry is foundational to our IDM 2.0 strategy and it’s well underway.
Today, measured with our lifetime deal value of greater than $15 billion, greater than $15 billion of revenue by 2030 per year, and this is based on leadership transistor technologies. Critically underpinned by our advanced packaging, our wafer-level packaging that will allow us to also leverage our older fab assets in powerful and exciting ways, and our greater mature node portfolio with UMC and Tower.
And the 18A design wins with our client and data center are already well underway, literally at scale capabilities coming into 18A. And we already have 5 customer commits and almost 50 test chips underway for 18A. Literally, we’ve captured the industry’s attention. Revenue leverages our at-scale existing infrastructure and resources. And the Foundry will produce healthy margins that will accrete significantly to EPS and free cash flow over time.
This is one of my favorite slides from Intel’s Direct Connect event, and it shows the huge potential benefits of being the world’s only systems foundry for the AI era. And fundamentally, we got to be a good foundry. We have to have great wafers and packaging technologies. But as you see in this picture, it leverages everything that Intel does.
Clearly, our advanced packaging technologies, this pursuit of Moore’s Law in 3 dimensions: our software, our systems, our memory, our thermal and interconnect technologies, all of them being able to be leveraged by our foundry customers.
And clearly, this idea of our product teams derisking and driving our factory networks, enormous value to our external foundry customers. And then our external foundry customers are making our foundry teams better, driving their technology creation, which also helps our product teams. Clearly, this creates this powerful virtuous cycle on this AI-driven continuum that is a value driver for Intel Corporation. This is a powerful strategy for the future.
And with that, I’ll now hand it over to Dave. Dave?
David Zinsner
Thanks, Pat.
Let me add my welcome to everyone joining today. I’m particularly excited to talk about our new operating model in more detail, specifically to highlight how this change will help drive profitability improvements for both Intel Foundry and Intel Products as we drive the overall corporate P&L to our target of 60% gross margin and 40% operating margin. While the recast of our reporting segments does not change our consolidated results, it does provide significantly better transparency and accountability, which will lead to better decision-making and improved profitability over time.
Let me take a minute to walk through the change. Our prior operating model allocated the cost of Intel Foundry into the product businesses, obscuring the underlying economics of both. The new model creates a foundry relationship between our manufacturing assets and our product groups, whereby Intel Products will now purchase wafers and services from Intel Foundry at a fair market price.
The fair-market wafer price becomes a major revenue driver for Intel Foundry and shows up as cost for Intel Products. Similar to how fabless companies purchase wafers from their foundry partners, each group will then have the information they need to optimize cost and improve profitability.
To determine fair market pricing, we benchmarked to the industry, taking into consideration the relative competitiveness of our wafers at any given process node. We released the 8-K about an hour ago, including the recast operating segments for 2021, ’22, and ’23. On this page, we’re just showing the numbers for ’23.
You’ll note that the P&L for Intel Products is actually quite healthy. It’s the first time you have had the information needed for an apples-to-apples compare with external fabless peers. Now that Intel Products has better visibility into what is really driving their costs, we see a clear opportunity to drive those businesses collectively to approach 60% gross margin and 40% operating margin.
Looking at Intel Foundry, the day 1 P&L is obviously very challenged, but we see ample opportunity that I’ll walk through in detail to drive to breakeven over the next few years and ultimately to our target of approaching 40% gross and 30% operating margin by 2030. The last point I’d highlight on this page is the large elimination line.
As I stated earlier, the recast doesn’t change consolidated results, and the elimination effectively adjusts for the intercompany activity between Intel Foundry and Intel Products.
If you look at Intel Products, the 10% operating margin that we showed in the allocated cost model increases to 24% in the fair market price model, solidly profitable at this point in the cycle. In fact, despite a significant drop in revenue, operating margin was flat year-over-year in ’23, really a testament to both the austerity measures we took as well as the very early progress identifying structural cost savings in the new model.
There are a number of areas Intel Products will focus on to drive continued improvement. We had execution challenges, which impacted the competitiveness of our products. As we’ve improved our execution, we’re introducing better products that can command higher margins. In our new model, the product businesses also sharpen their focus on architectural choices to lower the cost.
In addition, we’ve already seen meaningful reductions in test times, and of course, we expect improved operating leverage as we return to growth. We’re obviously overspending right now as we invest in businesses that are nascent today but drive significant revenue growth over time. As these businesses grow, we expect meaningful operating expense leverage.
So now looking at Intel Foundry. I want to spend a few minutes talking about the different drivers to get the business to breakeven in the next few years and ultimately to our longer-term target approaching 40-30 by 2030.
There are 4 areas to drive profitability.
First is transistor leadership. As Pat mentioned earlier, our current mix of wafer starts is heavily skewed towards nodes where we are not competitive on transistor performance or cost. That all changes as we mix towards Intel 18A. We’ll see meaningful ASP uplift as well as a more competitive cost structure, which will drive improved profitability.
Second is internal mix and scale. An added benefit of improved transistor performance is increased demand for those transistors. Intel Foundry lost Intel Products tiles to external suppliers due to the process deficits we discussed. As Intel Foundry regains process leadership with 18A, it has already won future Intel Products tiles, which increase scale. We think there’s roughly 2 modes worth of volume that will move back into Intel Foundry.
And as Pat mentioned, we have a number of external foundry customers that we’ve signed up to drive increased scale for the business. As we increase scale, that improves utilization and ensures we can deliver a return on the capital we’re investing.
Another unappreciated aspect of Intel Foundry going forward is advanced packaging. Many don’t realize advanced packaging will drive significant wafer demand going forward due to the base die used in most advanced packaging. Third is capital and cost efficiency.
I’ll unpack some of the efficiencies we’re already seeing in a minute, but one of the more significant contributors that Pat mentioned was CHIPS, both in the U.S. and EU and the investment tax credit, which is otherwise known as AMIC. This will reduce capital spend, improve depreciation, and ultimately improve our cost structure.
I’d also note that we’re currently spending a lot to drive five-nodes-in-four-years. As I mentioned on the Q4 earnings call, this year will represent peak start-up costs for five-nodes-in-four-years. As we return to a more normal 2-year cadence of node transition, we’ll have natural profitability leverage in both our cost of goods sold and our OpEx, which really touches on the fourth bucket of OpEx leverage.
In addition to a more normal node cadence, we’re actively driving OpEx efficiencies to align to our profitability targets. Importantly, many of the drivers to breakeven are fully in our control and somewhat independent of revenue growth. Although it’s always important in a fixed cost business, faster revenue growth allows us to achieve our targets more quickly.
Today, we are multiples higher than our longer-term target of 10% OpEx to revenue, but the path to breakeven embeds a modest goal of 15% to 20% that we think we can beat. We believe there is meaningful value creation as we execute on this path to breakeven. Had Intel Foundry been breakeven in 2023, we would have added more than $1.40 to EPS, more than doubling our reported EPS for last year.
As we move from breakeven to our 2030 target, the buckets don’t change, but we do get the added benefit of incremental efficiencies that drive capital avoidance today but take time to show up in the P&L. To help drive this process, we hired Lorenzo Flores as CFO of Intel Foundry. As many of you know, Lorenzo was the CFO of Xilinx, and more recently, he was Vice Chair of Kioxia. This page lays out the priorities Lorenzo has as he comes aboard, and I’m excited to have him join.
Let me click in on some of our focus areas for efficiencies. Number one is expedites. Expedites negatively impact the efficiency of a wafer fab. Semiconductor manufacturing is an extremely precise process that requires wafers to move through the fab at very specific intervals. Expedites interrupt this process, which is why pure-play foundries charge a very steep premium and why fabless companies use them sparingly.
Our IDM 1.0 segmentation hid the true cost of expedites from the product teams. So they requested them as a common course of business, and our manufacturing teams put capital in place to accommodate that inherent inefficiency.
Since introducing market-based pricing last year, we’ve already seen a 95% reduction in expedites, with fair market pricing driving better decision-making. We’re also already seeing 5% to 10% improvement in capital per wafer start on new capacity, driving improved cost structure for the business.
I talked about product design decisions changing under the new segmentation, and we’re seeing this change manifested in a 75% reduction of test times for a next-generation client product, again driven by the introduction of market-based pricing. We’ve also realized a greater than 10% reduction in engineering sample requests as product teams see direct charges from Intel Foundry. It’s going to take time to see a lot of these benefits flow through the P&L. But we’re already seeing behavioral changes and broad visibility to the new P&L will only accelerate that progress.
So let’s look at our midpoint targets for the business. We expect margin improvement on the product side with leadership products and better OpEx leverage. As we reach breakeven for Intel Foundry driven by margin levers that are very much in our control, we should see consolidated business with gross margin well into the 50s and operating margin approaching 30%.
If you look at the long term, we think we can drive 60% gross margin and 40% operating margin by 2030. Intel Products should be in the high 50s gross margin, 30s operating margin, and Intel Foundry should be in the high 30s gross margin and high 20% operating margin.
There is massive value creation as we drive the business to breakeven for Intel Foundry and ultimately, for the company to 60% gross margin and 40% operating margin. As Pat mentioned, the transformation truly is well underway, and we’re excited to execute this for our owners.
And now let me head over to join Pat and John Pitzer, who will moderate the Q&A for us.
John Pitzer
Thanks, Dave.
We’re now going to transition to the Q&A portion of the webinar. [Operator Instructions] I would remind you per my introductory commentary that we are in our quiet period and we will report Q1 earnings after the close of market on April 25. As such, please keep your questions to the topic at hand today, the release of the 8-K and the webinar detail.
With that, I’ll ask Jonathan to give us our first question.
Question-and-Answer Session
Operator
And our first question comes from the line of Timothy Arcuri from UBS.
Timothy Arcuri
Dave, I’m just wondering if you can bridge this new model versus the prior stated model that was 60/40 that was on a corporate basis? Can you just give us a bridge sort of apples-to-apples what this new model maps back to the prior model? Is this better or is it worse?
David Zinsner
Just so I understand, Tim, you mean our goal for 2030 is 60/40, the same as what we had beforehand. Is that what you’re saying?
Timothy Arcuri
Yes. I mean, before you had on a consolidated basis, you were at 60-40, so I’m asking you to map the new model versus…
David Zinsner
Yes, this is basically a similar view as to how — what we looked — I mean, we’re just recutting the data so this is really a similar view. Obviously, we had a different trajectory back at the Investor Day, but obviously, the cycle hit, which delayed the drivers that got us at the 60/40 model. But we’ve been talking about 60/40 for a while. What we’re doing is we’re unpacking how we think we will get there from an Intel Foundry perspective and an Intel Products perspective.
John Pitzer
Tim, do you have a follow-up?
Timothy Arcuri
I do. Yes, Dave. So most of the op margin ramp is very back half loaded of this decade. Does it sort of imply that it’s more key to 14A and 10A versus 18A? Can you kind of talk about sort of what the keys are, if you can map it back to the process nodes? Is it really something gets more of a 14A and key versus an 18A and key?
Pat Gelsinger
Let me actually start on that one, Tim, and then I’ll have Dave give some of the math behind it. What happens is in ’25, we’ll start ramping Intel 3. In 26, we’ll start ramping 18A. And when we get to 18A, we believe that we have competitiveness and the cost structure. The pricing will be based on market pricing, and we have a very effective price performance metrics associated with it.
But the ramp of that into the product line and somewhat as you saw in the slides really only hits the large volumes in ’27 and ’28. So you see it just naturally rolling into the business as we roll off Intel 10, which had a very ineffective both power performance as well as cost structure. Obviously, 14A makes that better but 14A doesn’t even qualify until 2027, and its ramp starts really hitting the P&L in ’28. So all of these sort of cascade into the business. And unfortunately in the semiconductor business, you have this ramp timing associated with it being the overall portion of the business model.
And obviously, when you look back at some of those graphs as well, you see the effects of the mature nodes tapering down, the advanced packaging, which doesn’t have the same level of margin structure. But taking all of that together, we’re just on track to deliver. And it really is the defining point, as I’ve called it, the pre-EUV, and the post-EUV for Intel. In the post-EUV era, we see that we’re very competitive now in price performance back to leadership. And the pre-EUV era, we’ve carried a lot of cost and uncompetitiveness which essentially had us pricing those wafers very low, right?
And we see this post-EUV era as one that very few companies are going to be able to achieved. So we see it as a very, I’ll say, scarcity value in that period of time. So it just layers through on track for all of the node transitions, product commitments and the bulk of those wafers are Intel.
But then in the back end of the period, we also start to see the external foundry wafer start to layer in as well. And all of those start to offset the Smart Capital, the benefits that we see from that in terms of the cost structure, competitiveness globally. And all of those things taken together are what put us on this trajectory through the decade. Dave?
David Zinsner
The only other thing I’d add is, roughly, when we get to breakeven, we’re about halfway there from a total company perspective in terms of our operating margin goals. So it’s not like it’s all back-end loaded. It does show up meaningfully by the midpoint. And then, of course, we’ll see more of that to drive it to 40% operating margins. What is true, though, is ’24 and ’25 are probably a little bit more slower in terms of the improvement.
Obviously, ’24, we expect to be the trough on the Intel Foundry side, and that’s going to limit the amount of operating margin expansion we have this year. And next year is kind of continuation of some of that start-up costs as we kind of work through the five-nodes-in-four-years. But once we get out of that period, a lot of the things that Pat just talked about will really start showing up in the P&L.
Operator
Our next question comes from the line of Joe Moore from Morgan Stanley.
Joe Moore
I wonder as you built this segmentation where you allocated some of the costs around process technology development and things like that, is that all being borne by the IFS business or is some of that share? And is this a structure that are these really like 2 separate organizations where the costs are allocated as they would be if it were separated that way?
Pat Gelsinger
Yes. And starting at the highest level, we’re treating Intel Foundry like a foundry, meaning that it’s carrying the capital cost, the depreciation cost, all of the technology development and R&D costs, and it is buying wafers at competitive pricing to the industry. So that is the high-level thesis behind this operating model, treating Intel Foundry like a proper foundry to a fabless semiconductor, Intel Products. And that’s the separation at the highest level.
We based that wafer pricing, test pricing, packaging, pricing on the established pricing that is visible in the industry. And those are reasonably well known, and we benchmarked ourselves very thoroughly against those to establish this pricing model. And that’s why we say, it brings proper discipline to the Intel Products group because now they’re looking at every one of these things that they weren’t necessarily being managed or controlled on before.
And of course, the Intel Foundry is a proper foundry. And it is building everything associated with the technologies, the technology road map to meet the internal but also the external foundry requirements of the industry.
Taken together, we’re building the clean legal separation of the Intel Foundry. We’re rebuilding the ERP systems of the company, establishing those separate data flows that external customers can have their confidential information protected and their capacity corridors guaranteed as well. So all of this is a proper foundry with a proper fabless business with it.
John Pitzer
Joe, do you have a follow-up?
Joe Moore
Yes. And just in terms of thinking about the kind of sum of the parts, would I assume that the book value of the company would mostly kind of accrue to the IFS side? I know you’re not segmenting it that way, but how should I sort of think about where the balance sheet value would go?
David Zinsner
Yes, I think that’s right. I mean, the biggest part of our net book value is obviously our PP&E, and most of that shows up in the foundry side. So I would say within a couple of years, that’s like a $100 billion number net book value for just Intel Foundry. And I think your knock-on question here might be, well, how do you look at that in terms of valuation?
And even the worst foundries are getting 2x net book value and the best, of course, is getting 5-plus net book value So this is where we think a lot of the value unlock in the business will come from is that net book value generating a good return.
Pat Gelsinger
Yes. And this also gives them a very clear view of the sustained profitability of a very healthy Intel Products margin business and the growth characteristics of that. So it presents those in I’ll say a very proper market basis as you start looking at the overall sum of the parts valuation for Intel over time.
Operator
And our next question comes from the line of Vivek Arya from Bank of America Securities.
Vivek Arya
I’m curious, for the first one, what are the starting gross margins for your Product, Foundry and others so just we have a baseline to compare to your longer-term targets? And also, David, if you wouldn’t mind giving what the free cash flow targets are.
David Zinsner
So just — well, we went through — that’s part of the 8-K actually is what the gross margin or what the operating margins are. Were you looking for what the gross margins looked like? Is that…
Vivek Arya
That’s right, gross margins for Products, Foundry and Others so as a baseline, right, because otherwise, how do we compare where you’re doing?
David Zinsner
Yes. I mean, maybe just rough order, we’re not really going through the details of that. But I’d say rough order, you’re talking about a 50%, plus or minus, business on the Products side. And on the Foundry side, it’s a negative gross margin. And Pat showed a slide that showed wafer margins at 0% at the kind of Intel 7 level or pre-EUV level. And once you have the start-up costs, that drives a negative gross margin. So lots of opportunity to expand those to get to the targets.
John Pitzer
Vivek, do you have a follow-up?
Vivek Arya
Yes. So Pat, the last time you had an Analyst Day, I think you gave targets out for 2026, and now you are giving targets out to 2030. So my question is, what are the big differences in the market you saw when you gave the ’26 targets versus what you’re seeing in kind of pushing them out to 2030? What are the pluses and minuses in your market assumptions, right, with the new targets versus what you had set before?
Pat Gelsinger
Yes. Thank you, Vivek. And if we build up from a process technology perspective, five-nodes-in-four-years, the costs associated with those to build it and the pathway to profitability, I’ll say essentially no change from what we spoke about in our Investor Day. From getting back to product competitiveness, we’re on track to go deliver that. So again, the core products, approximately where we were before.
The couple of major changes from the headwinds that we saw was clearly, we weren’t expecting some of the market cycles that we went through in the post-COVID era, things like a stronger PC market growth was expected. We also expected that we would have been more successful with our accelerator share gains and where we would be by this point in time. So those would be, I say, the 2 biggest headwinds that weren’t accounted for before.
What we’ve laid out today is we’ve tried to take very reasonable, modest assumptions for our core business where we’re sort of in the core area, in the mid- to low single digits growth rate. We add on top of that some share gains for accelerators over the periods. Additionally, some improvements or additional revenue growth for our Foundry business that we’re seeing as we look over that cycle as well, leading us to the mid- to upper single digits for an aggregate growth rate in the modeling that we’ve seen here.
Obviously, we see plenty of opportunity to beat those as we have built this modeling. We do think that the scarcity value, the deal value that we’re building up with Foundry, the improving competitiveness and categories like AIPC will drive growth. But we’ve taken very reasonable assumptions throughout the modeling that we’ve shown you today.
And we believe the segmentation will drive the right business characteristics by our internal teams to improve. We’ve also made major steps with things like Mobileye, Altera, IMS, and their segmentation and presenting those to the marketplace.
And taken all together, we believe today’s model is something that in this market environment that we’re well underway, our Intel Foundry revenue is a great opportunity for growth, and the Intel Products is a very healthy, fabless business that has good growth potential to it as well.
Operator
And our next question comes from the line of Aaron Rakers from Wells Fargo.
Aaron Rakers
I guess I want to kind of dig a little bit deeper into the product operating margin trajectory here as you look forward. Looking at the recast segment of the Data Center and AI piece of business, I know that was about a little less than 13% operating margin this last calendar year. Dave, how do you see — what’s the right normalization of the operating profile in that business? When I look back, I think back in ’21, it was implied at close to 44%. How do you think about the trajectory of that?
David Zinsner
Yes. And I think you’re dead on, on where the trajectory is. That is kind of the level of operating margin you should see in one of those businesses, a higher grower business — higher-growing business with significant barriers to entry and good margin profile. So that would be the goal for sure. Obviously, we’re investing a lot right now, particularly in the area of accelerators in businesses that have negligible revenue.
And as those businesses become meaningful and grow at a rapid rate, we should get good OpEx leverage. I mean, I really feel like a lot of where we’re spending in terms of OpEx right now is actually where we can kind of sustain it over time. So there might be areas where we may invest more in but I think there are opportunities to drive more efficiency as well. So there’s a tremendous opportunity for us as the businesses grow to get significant leverage.
Pat Gelsinger
And maybe just to add on to that a little bit more, particularly in the data center, this is the first area that we’ll be seeing the benefits of Intel 3. And as we bring Granite Rapids and Sierra Forest to market now we’ve gone, for the first time in many years, having a process technology leadership with great new products as well. So this is going to start stabilizing our market share position.
Also, the Intel 3 from a foundry perspective has much better margins and competitiveness associated with it. So it’s sort of the first moment that you see Intel Foundry contributing in Intel Products. Those products only launch this year so they’ll be ramping mostly, you’ll see the benefits of that starting to hit the P&L next year. And obviously, as we go to Clearwater Forest, our next year product, a major leap ahead with 18A, both better cost structure from foundry, better competitiveness as well.
And then as Dave said, of course, very important is starting to bring our accelerators to the marketplace, and we’ll be updating you on things like Gaudi, Falcon Shores, in the near future. So all of those, we’re investing heavily. We have to get that business back to the profitability and operating margin that you and we would expect from it.
John Pitzer
Aaron, do you have a follow-up question?
Aaron Rakers
Yes, I do, John, and I appreciate that. So I guess I wanted to ask on just the IFS business now that we unpacked this and you look at the P&L, I’m curious, as you change the cadence beyond these five-nodes-in-four-years and you go to back to a normalized cadence, how do you guys depreciate the equipment in IFS? Or is there any, at the longer end of that horizon, the opportunity to extend the depreciation cycle of the equipment leverage in IFS to kind of drive gross margin?
David Zinsner
Yes, it’s a good question. I mean, in reality, we’ve moved the depreciation to 8 years. I wouldn’t envision us changing that going forward. Of course, we’ll evaluate it on an ongoing basis. But there’s no doubt that we will get longer use of the assets. Might not be reflected in the depreciation schedule, but it will be reflected in the return on invested capital and the return on assets. So that, at the end of the day, is really what matters and that for sure is going to be a dynamic that we’ll see.
Pat Gelsinger
Yes. And we’re going to put a lot of emphasis on extending the life. And if you go back to that little cartoon picture of nodes dropping off very quickly versus long-life nodes, this becomes very important for the overall business model. It’s something that has really been the long line free cash flow portion of the foundry assets.
And for us, when we get over the EUV hurdle, we have the most modern, productive EUV fleet available in the industry. We also see that we have good equipment reuse between the EUV nodes. Also as we move to this 3D construct base die will extend the life of processes like Intel 3, we’re also doing more node variations, as you saw in the road map that I presented. Also advanced packaging will extend the life of the capital assets and then relationships like we’ve described with UMC.
So all of these taken together become very rich portions of that improving ROIC as we start to build that long corpus of manufacturing capabilities that we have. That’s why I really define the world as pre-EUV, post-EUV. Very few are going to be there. There’s going to be scarcity value associated with those assets and capabilities. Also, one of the things I highlighted was that the cost per transistor, if you measure the cost per 100 million transistors, essentially pre-EUV to post-EUV, it’s about half, meaning it’s uncompetitive to be building chips on pre-EUV nodes going forward.
So all of these, we believe, drive a very effective capital assets. We’ve levered Smart Capital very effectively, very unique turf, and the AI-driven era will be putting significant value in driving those nodes for longer and longer in the business model that we’ve laid out.
Operator
Our next question comes from the line of Vijay Rakesh from Mizuho.
Vijay Rakesh
Just a quick question. I was wondering on some of the charts that you showed, you had the Foundry revenue going out to 2027, 2030. Just wondering what are your expectations were, actually out that margin profile for the Foundry business, what are the revenue expectations there?
Pat Gelsinger
Revenue expectations. Let me do revenue, and I’ll let Dave talk about margin. Revenue expectations, what we said is, today, it’s very small in one of the graphs, it ramps very smoothly as we go through the period. We expect we hit greater than $15 billion in revenue by the end of the period. And obviously, as you look at the impendent and then consolidated financials, we have more than double that associated with the revenues from the internal foundry.
And as we said, number 2 foundry by the end of the decade based solely on external revenue. But taken together, this is clearly the number 2 foundry in the industry. And that will be — the biggest portion of that is the EUV wafers or leading-edge nodes. But advanced packaging is a very healthy portion of that and then some amount of legacy nodes as well that were driving extending capital use associated with that. Dave, the margins?
David Zinsner
Yes. On the margin front, we’d expect the gross margin to be very similar to what we talked about for the model, for the overall Intel Foundry of 40%. We see that from the wafer perspective for sure. And we also see that in the packaging side. We think we will get very rich margins on a relative basis, given our technology and what we’re bringing to bear on advanced packaging.
Actually, when you look at it from a wafer margin perspective, it should be higher than that. But the expectation is that we will always be having some amount of start-up costs associated with the business that will kind of lock that into the 40%.
Pat Gelsinger
And given superior cost structures, as we’ve described, obviously, the margin structures that have been defined by the competitive environment, we think we have a very healthy umbrella to bring these wafer pricing into the marketplace to be able to both win share and improve margins over the cycle.
And as we deliver on leadership technology, they’ll command leadership pricing as well. So overall, we feel very confident that picture that we’ve laid out is one that we’ve taken conservative modeling of it, and we see plenty of opportunity to meet and beat it.
John Pitzer
Vijay, you have a follow-up, please?
Vijay Rakesh
Yes. Just on the Intel Products side, actually look out, is that a mix of in-sourcing versus an outsourced foundry that you are embedding in the Intel Products margin or profitability target as you look out to 2030?
Pat Gelsinger
Yes. And we do expect that we’re, as we’ve said, bringing some wafers home today, we’re sort of peaking right now in this period of time as we’ve worked through our stumbles as a company where we’ve leveraged external foundries most aggressively. Now we fully expect that part of our Smart Capital strategy is that we’ll keep using external foundries over time, even though those will be at a lesser scale than we’re using today.
So we do fully expect that those foundry relationships, super important for us in the future, but they’ll play a less critical or less substantial portion of the overall P&L as we go through the period. So peaking this year and next year and then moderating over time. But we’ve built into the margin assumptions for the product groups some continued use of external foundries. And part of that use is for some specific nodes that we don’t necessarily have for some capacity balancing.
And in the cases they’re just capable of things that we haven’t necessarily yet added through our foundry portfolio. These are important relationships for Intel and part of being in Intel Products. And then Intel Foundry, the products team will take continued use of them for as far as we can see into the future. That’s part of our business plan and built into the P&L modeling that we showed you today.
Operator
And our next question comes from the line of Srini Pajjuri from Raymond James.
Srini Pajjuri
Pat, just to follow up to the previous question. Could you remind us what percent of your revenue is outsourced to external foundries today? And also, I guess to what extent is your foundry breakeven dependent on in-sourcing some of the volume? Because I think Dave, you said that you can control a lot of the things to bring the foundry business to breakeven. We’re just trying to understand what role that in-sourcing will play in achieving the breakeven in your Foundry business.
Pat Gelsinger
Yes. So it’s on the order of 30-ish percent of our wafers today that we bring in externally. We’ll be in-sourcing some level. As I said, a couple of fab modules, we expect over this period of time. So we expect that to moderate to down below 20% over the period even as we continue to use. As I answered the last question, external foundry is an important part of our business strategy, but we will be bringing more of those wafers home.
That helps us in a number of dimensions in terms of cost, obviously, the consolidation benefits that we get from that. It also will allow us to extend the life of some of the factory nodes as well, so another important point of building that out. And of course, as we’ve seen in different supply-demand scenarios, the ability to control that supply is another benefit of the business model. So all of that taken together, a good balanced outlook as we moderate the external foundries.
And we appreciate the unique relationships that we have to help them tell a very important period of time. Where we weren’t competitive, we’ve leveraged them and as we look to the future, we continue to invest in key partnerships that we have with the foundry industry.
David Zinsner
The only other thing I’d add is just keep in mind that we have pretty good line of sight because the designs are already going on and have gone on that bring those wafers back at 18A and beyond. So I think we have a good sense of what tiles will come back into the fab, which is — and I think we’ve modeled it conservatively when you look at what Pat showed. So I think this is very much a doable model for us.
John Pitzer
Srini, do you have a follow-up?
Srini Pajjuri
Yes, John, thank you. Pat, just a clarification on something you said. You mentioned that 2026 is when you expect 18A volume and 2025 interim grid volume. Is that I’m just — I thought you previously said 2025 is when 18A will ramp. So I’m just trying to make sure that this is — there’s no change to your view as to when 18A will ramp to volume.
Pat Gelsinger
Yes, thank you for that clarifying question. 18A goes into volume ramp in 2025. However, the volumes that you’re seeing and the impact of it on the P&L are pretty modest until you get to 2026. So the question was really asking, shouldn’t we be seeing the margins improve more rapidly? And it really is the bulk of wafers in 2025 are still Intel 3. It was a bit bigger, but still the bulk of our wafers in 2025 are driven by Intel 7 and Intel 10. So that moderates the margin benefits that we get as we ramp the new EUV node.
So we’ll see a good amount of Intel 3 in 2025, a small amount of 18A wafers. We’ll see a good amount of 18A wafers in 2026. And obviously, those volumes just continue to swing to the more modern post-EUV nodes over the horizon for it. So no change in the 18A schedules, goes to production release this year. We’ll be ramping first products, Panther Lake, Clearwater Forest. The first desktop product or the first client product and the first server product in 2025 will start to ramp, and it will be high volumes in 2026 for 18A.
David Zinsner
And we just reviewed the Clearwater Forest today and looking good.
Pat Gelsinger
Everything is looking good on that journey. And of course, we’ll start to layer in the external foundry customers on Intel 3 and 18A. They will follow typically the Intel products by a couple of quarters, but we’re seeing a very healthy pipeline of those activities as we indicated early. So everything there exactly on schedule.
And I’ll just say, this is a key piece. Why did we come to this discussion today? Well, part of it was we had to get line of sight on getting back to technology leadership. Now that, that’s firmly in view of how we’re going to finish that work that we laid out in five-nodes-in-four-years, now we have to have the business and financial model and the organizational structures to fully realize the vision that we laid out. That’s why we’re here today. Everything is on track.
John Pitzer
Jonathan, we have time for one last question, please.
Operator
And our final question for today comes from the line of C.J. Muse from Cantor Fitzgerald.
C.J. Muse
I guess first question, hoping you can help me reconcile where Intel Products can kind of achieve where the foundry versus the Intel Foundry side, it has to make obviously a multiyear investment decision. So I guess the clear question is that how early does Intel products need to commit to capacity at Intel Foundry?
Pat Gelsinger
Yes. And I’d say it’s not, I’ll say, atypical from how any foundry would be chosen by a fabless provider. Now obviously, in this case, C.J., that the critical aspects, I’m also driving our internal product teams to be customer zero, where we’re committing that first design. That’s a deep collaboration of technology, of the early EDA and IP, and that’s part of what makes this such a powerful model is we are driving those early designs very aggressively into Intel Foundry.
So it’s earlier in that sense that might be the case for other foundry alternatives externally. But the bulk of their wafers are not those leading-edge very first designs but over the life of it. And in those cases, it’s looking at the business benefits of that. We said we’re pricing our wafers with competitive technology below market prices so that we win in this business, so that presents a business advantage to the Intel Products.
And clearly, the revenues, the volumes, the factory builds, this is essentially a capacity corridor commitment and conversation that we’re having between Intel Products and Intel Foundry. And as we’re teasing these apart, we’re building those new muscles of how Intel Products works with a proper foundry company as well. We’re rebuilding the ERP systems for the company, the IP management. All of these are things that are now well underway as a company.
So we’re quite proud of the progress that we’ve made for a world-class Intel Foundry and a world-class Intel Products all as part of the Intel family.
John Pitzer
C.J., do you have a follow-up?
C.J. Muse
Yes, quick one. You showed us $19 billion in Foundry revenues pro forma from 2023. Curious to get to breakeven and then to get to the target operating model, what kind of revenues in total should we be thinking about? And is there an ideal percentage, internal versus external, at those numbers?
David Zinsner
Yes, maybe I’ll start and yes. I mean, I would say at the high level, C.J., we don’t need a lot of revenue growth to drive this. The underlying improvement of moving to 18A, the margins of 18A, both in terms of pricing and in terms of cost structure, the move of the tiles as we do that, the efficiencies we think we’re already starting to see the benefits of, more to come in that regard. All of those things really aren’t a function of volume increases, and we think we can get to breakeven with very, very modest rate.
Now that said, there — we do — we have $15 billion of lifetime deal value. As Pat said, we do believe we’re going to be growing that over time. We do intend to be the number 2 foundry player by 2030. So we are going to get growth but we don’t need it necessarily to get to the breakeven target.
Pat Gelsinger
Yes. As you get to 2030, I mean, we’re still expecting that 2/3, 70% of our wafers are the internal products that we’ve built up, that 30-ish percent being external wafers by that point in time. And there’s pluses and minuses around that. My objective, though, is to grow the external business as fast as we can and win as many of the next-generation designs to the Intel Foundry.
Obviously, I have to earn our customers’ trust. We have to imbue in them the confidence that they can commit their business to Intel and have that ability to drive their business through the Intel Foundry, and they have come to have a level of expectations of what that means for very good foundry alternatives in the industry, and we have to become one of those very good foundry. But we think the assumptions are very reasonable, the relationships that we’re seeing.
And the energy that we saw coming out of Direct Connect that we held a couple of weeks ago was really profound that we saw the EDA, the IP momentum that we saw in the industry. You heard testimonials coming from people like Rene Haas from ARM, the great results that they’re already seeing on 18A, world-class PPA and area results that they are realizing today, the full cadre of EDA support, all these pieces are coming together very profoundly.
So our objective is be the foundry of choice for all of those suppliers and particularly driven by this AI era. We see a lot of momentum beginning with packaging as the on-ramp to build that trust the momentum with the industry.
So maybe in wrapping up our time today. Hey, we appreciate the time of all of you, our key stakeholders. And today is the next step in the IDM 2.0 journey, and I feel like we’ve been preparing for today’s call for the last 3 years. Our founders, Noyce, Moore, Grove created the company that measures what matters. If you don’t measure it, you can’t manage it. And our updated segments that we’ve given you today reflects this mentality exactly.
In summary, Intel is going to build leading-edge products and the world’s only Western at-scale sustainable leading-edge systems foundry. And I should be clear from today’s discussion, we expect significant financial upside as we progress toward those targets, the financial models we described to you today. And as our founders told us, we will not be encumbered by history. We’re going to go out and do something wonderful.
Thank you so much for joining us today.
David Zinsner
Thanks.
Operator
Thank you, ladies and gentlemen, for your participation in today’s conference. This does conclude the program. You may now disconnect. Good day.
Read the full article here