管理層發言
Thank you for standing by, and welcome to Intel Corporation's Second Quarter 2026 Earnings Conference Call. 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, Vice President, Investor Relations. Please go ahead, sir.
Thank you, Jonathan, and good afternoon to everyone joining us today. By now, you should have received a copy of the Q2 earnings release and presentation. Both are available on our Investor Relations website, intc.com. For those joining us online today, this presentation is also available on our webcast window. I am joined today by our CEO, Lip-Bu Tan; and our CFO, David Zinsner. Lip-Bu will open up with comments on second quarter results and update the progress we're making on strategic priorities. Dave will then discuss our overall financial results, including third quarter guidance before we transition to answer your questions. Before we begin, please note that today's presentation does contain forward-looking statements based on the environment as we currently see it. As such, they are subject to various risks and uncertainties. It also contains reference to non-GAAP financial measures that we believe provide useful information to our investors. Our earnings release, most recent annual report on Form 10-K and other filings with the SEC provide more information on specific risk factors that could cause the actual results to differ materially from our expectations. They also provide additional information on our non-GAAP financial measures, including reconciliation where appropriate to our corresponding GAAP financial measures. With that, let me turn it over to Lip-Bu.
Thank you, John, and good afternoon, everyone. Q2 was another quarter of solid execution. Revenue, gross margin, earnings per share were above our guidance. This marks the seventh consecutive quarter of exceeding our financial expectations. Our core message is simple. Strong demand for our products continues to outpace our growing supply. Our design, manufacturing, execution is improving and the operating discipline we put in place 15 months ago is beginning to show tangible results. Today, we are seeing the strongest revenue growth in more than 15 years. Our cultural transformation continues and our organization is already operating with greater efficiency, moving faster, making better decisions and staying closer to customers. Our recent announcement to deepen our collaboration with Google Cloud will help to accelerate the transformation as we fully embrace an AI-first mentality throughout operations. We also continue to strengthen our leadership team with world-class talent. The surging demand and rapid build-out of compute infrastructure across the world creates a meaningful opportunity for us in our product business as well as our foundry business. The industry is facing one of the most severe supply constraints in its history across leading-edge logic silicon wafers, memory and substrates. These shortages will persist for the foreseeable future. Intel is well positioned to benefit from this strong, sustained demand with three strategically important assets: our x86 CPU franchise, our advanced packaging technology and our vast wafer foundry network. As AI expands from training to inference and increasingly to agentic and multi-agent systems, general purpose server CPU density continues to increase, and our core server CPU franchise is growing faster than ever. Demand signals from our customers are driving increased confidence. And as Dave will discuss in more detail, we are substantially increasing our investments to support this improving demand outlook. Let me talk first about Intel Foundry. My confidence in our foundry process roadmap has grown significantly since joining over a year ago. I am more confident than ever of the strategic and significant unique value proposition of Intel Foundry. During Q2, our factories across Intel 7, Intel 3 and Intel 18A exceeded internal volume targets, driven by improving yields, better cycle times and increasing wafer starts. 18A output increased meaningfully in the quarter. Yields continue to track ahead of expectations. We are now ramping multiple new products on 18A, while supporting growing demand for our lead products, including Panther Lake and Wildcat Lake. I keep raising the bar on the internal targets, and the team continues to meet the challenge. The successful volume ramp of 18A for our internal products provides important validation as Intel Foundry engages with external customers. We also began risk production of 18A-P, providing additional performance and power advantages, while maintaining IP and design compatibility with Intel 18A, positioning 18A-P as a competitive node for external customers. Looking beyond 18A, I'm encouraged by our progress on Intel 14A. Defect density and transistor performance are all outpacing 18A development. PDK 0.5 is now complete, and PDK 0.9 is on track for October. We continue to build out and validate the IP portfolio for 14A as we position the 14A family for broad-based adoption across a wide range of customers. I'm pleased to see the increasing momentum on customer engagements for Intel 14A, and I'm increasingly confident that 14A will be a highly competitive process offering across key vectors of performance, power, density, cost and schedule. With encouraging external customer progress and increased demand for our internal products, we remain on track for 14A risk production for our internal products in the second half of 2027, and we made the decision in Q2 to be fully committed to high-volume ramp in 2028. Lastly, on advanced packaging, customer interest for EMIB-T continues to be very high. The technology is compelling, providing capabilities for advanced AI silicon solutions that are not possible with today's mainstream offerings. We continue to have a growing EMIB-T backlog; yield and reliability are hitting targets, and we are focused on ramping the technology into high volume and high quality to support customer ramps in 2027. Turning to Intel products. We recently renamed our PC business to our Client Computing and Physical AI Group, or CCPG. We did this to recognize the growing opportunity for AI at the edge. I'm excited to have new, strengthened leadership driving this effort. In our core PC client segment, Intel 18A is now in volume production across multiple commercial and consumer products. Our factory output continues to increase sequentially every month. The successful high-volume ramp of 18A for our internal products provides important validation as Intel Foundry engages with external customers. We still have work to do to establish a strong footprint in the edge and physical AI ecosystem, but we see this opportunity as an important future growth driver. Our Data Center and AI Group delivered a solid quarter. Demand accelerated across cloud and enterprise as customers increasingly recognize the critical role that CPUs in general and x86 CPUs, in particular, play in AI infrastructure. Q2 year-over-year server growth was the strongest on record, and Xeon 6 continues to be one of the fastest ramping products in Intel history, reflecting improving execution and strong customer demand. We have also strengthened our outlook with additional strategic customer wins and long-term agreements in Q2. Our top priority is to ramp capacity and improve factory output as quickly as possible to support customer demand even as we work to improve our competitive roadmap. We also extended our heterogeneous AI strategy through a multiyear collaboration with SambaNova. We are pleased with their growing momentum as we work with them to drive performance and power improvements with disaggregated inference. Lastly, we continue to make steady progress in our newly announced design services business with revenue growing nearly 3x year-over-year. We see tremendous opportunities to leverage our strong x86-based general purpose computing franchise to build more purpose-built computing products for the AI era. Our unique assets in end-to-end design, our IP portfolio, combined with access to our leading-edge wafer and packaging capabilities, position us well to win in this fast-growing space. We are already making great strides in expanding our purpose-built portfolio from networking to compute and eventually accelerators. Our recently announced collaboration with Fortinet for their security processor is a strong step forward in our ASIC strategy in the quarter. As we look ahead, I'm excited to see the new Intel take shape. We are operating with greater speed, accountability and customer focus. While there's still significant work ahead, our priorities are clear: leverage our x86 computing franchise to strengthen our product leadership and establish Intel Foundry as a world-class wafer and packaging foundry business. Intel is uniquely positioned to benefit from the overwhelming demand for compute as the entire industry continues a rapid and sustained build-out of compute infrastructure. We are the only company that can design, manufacture, and build the entire range of computing solutions from general purpose, traditional CPUs and GPUs to more purpose-built ASICs and CPUs optimized for agentic AI, as we increasingly move from compute dominated by system on chip towards system in package. Our advanced packaging and wafer foundry capabilities become increasingly vital assets. Our strategy is clear and the pace of execution is accelerating. Opportunities in front of us are substantial. Our strategy is showing early results, and I'm confident that Intel is well positioned to help define the next era of computing. I want to thank our employees around the world for their focus, discipline and hard work every day and also thank our many customers, partners and suppliers for their continued trust in Intel. With that, let me turn the call over to Dave to walk through our financial results in more detail.
Thank you, Lip-Bu. We delivered another strong quarter, driven by robust demand and disciplined execution that resulted in upside to our supply. Second quarter revenue was $16.1 billion, $1.8 billion above the midpoint of our guide. And collectively, our AI-driven businesses grew greater than 70% year-over-year, including record data center growth and contributed approximately 70% of revenue. It's important to note that despite exceeding our expectations for wafer outs in the quarter, strengthening demand continues to outstrip our growing supply. Q2 non-GAAP gross margin was 41.8%, approximately 280 basis points better than guidance. The upside was driven by higher revenue, better yields and higher ASPs due to mix and pricing actions. We delivered second quarter non-GAAP earnings per share of $0.42 versus our guidance of $0.20 on higher revenue, stronger gross margins and solid operating leverage. Q2 operating cash flow was $7 billion, and we exited the quarter in a strong liquidity position, including approximately $30 billion in cash and short-term investments. Our Q2 results reflect the ongoing progress in our operational transformation. We're moving faster, holding ourselves more accountable and staying closer to our customers. Moving to segment results. CCPG revenue was $8.9 billion, up 15% sequentially and better than our expectations. The client TAM continued to hold up well despite broad component constraints and price inflation. Our AI PC revenue grew 26% sequentially and now represents two-thirds of our client revenue mix. In addition, we saw solid performance in our edge deployments, now representing roughly 10% of CCPG revenue. Operating profit for CCPG was $2.3 billion, 26% of revenue and down approximately $173 million quarter-over-quarter due to inventory charges taken to optimize our factory network to overall customer demand across client and server. Our Client Group has now brought 18A to full scale with 400-plus designs for Series 3 across consumer and commercial. During a period of inflationary pressures, CCPG brought Core Series 3 to market at the ideal time on its A stepping to provide cost-optimized mainstream compute capabilities. Our integrated Arc Graphics solution continues to see strong market adoption with 40-plus Arc integrated graphics designs across creator, workstation, commercial and gaming targeted designs. Building on our success in gaming notebooks in Q2, CCPG also introduced Intel Arc G-Series processors, a new family of products designed for next-generation handheld gaming systems, an additional vector for growth. On the commercial side, activations for our market-leading vPro manageability software have surged 1,500% over the last four quarters, underpinning that manageability and enhanced security are critical must-haves in the agentic workplace. We expect enterprise adoption of AI to be a long-term tailwind for CCPG, but our AI-driven market prospects don't stop there as the edge and physical AI opportunity is likely to at least match the client TAM over time. CCPG showcased this growing opportunity with 130 Series 3 design wins for edge AI applications, including brain and control deployments for robotics. DCAI revenue was $6.3 billion, an increase of 24% sequentially and 59% year-over-year, meaningfully ahead of expectations. The result was driven by strong demand across hyperscale and enterprise. We also continue to see strong momentum in our purpose-built silicon product line, with revenue up roughly 20% sequentially and nearly tripling year-over-year. Operating profit for DCAI was $2.5 billion, 40% of revenue and up approximately $1 billion quarter-over-quarter on higher revenue, improved product margins and lower operating expenses. Within the quarter, DCAI launched Xeon 6+ codenamed Clearwater Forest, our first server class product on 18A. The team announced rack scale and disaggregated inference innovations with partners SambaNova and Foxconn. In addition, DCAI further enhanced our connectivity offerings by introducing new controller and adapter products supporting data center, enterprise and telco applications, which scale from 100 to 200 gigabit Ethernet. Turning to Intel Foundry, revenue of $5.8 billion was up 6% sequentially on higher fab volumes driven by strong growth in Intel 18A with output approximately 25% above target and up more than 50% quarter-over-quarter. External foundry revenue was $293 million in the quarter. Intel Foundry operating loss in Q2 was $2.1 billion and $348 million better quarter-over-quarter as higher yields, improved cycle times and increased factory scale across Intel 4, 3 and 18A drove improved wafer costs. Progress on 18A has been very good. Intel Foundry has driven down the cost of our primary Panther Lake SKU by roughly 50% year-to-date and is on track for an additional 20% this year with further meaningful reductions planned in 2027. Within the quarter, in addition to delivering output above our expectation, Intel Foundry also entered risk production for 18A-P and met critical milestones towards delivering the 0.9 PDK for Intel 14A in October. We stepped up investments in Q2 for Intel 14A to prepare for risk production in 2027 and committed to high-volume ramps in 2028. Now turning to guidance. Looking ahead, customers continue to signal a strong and sustainable spending environment driven by the unprecedented demand for AI compute. Industry-wide supply constraints across wafers, memory and substrates remain the dominant challenge our customers are facing to support the AI infrastructure build-out. Our wafer output across our major nodes exceeded expectations from 90 days ago, and Q3 quarter-to-date 18A yields are trending ahead of targets set in March. Even with this strong execution and the positive trends as we enter the quarter, supply remains very tight and the near-term linearity of our supply growth is more skewed towards the end of Q3 and into Q4, especially for servers. From an end market perspective, we expect PC consumption to be subseasonal in the second half of the year and down low double digits percent for all of 2026, impacted by rising memory prices and constraints. This is in line with industry peers and third-party estimates. At the same time, improving supply, a strengthening product portfolio and encouraging tailwinds for edge deployments provide us with some positive offsets. Our outlook for server CPU demand has improved again since our last earnings report and we're forecasting strong double-digit unit growth for the industry this year and next, with momentum extending into 2028. Taken together, we're guiding Q3 revenue to a range of $15.8 billion to $16.8 billion. At the midpoint of $16.3 billion, we forecast gross margin of 42%, a tax rate of 11% and EPS of $0.38, all on a non-GAAP basis. We continue to tightly manage non-GAAP operating expenses to roughly $16.5 billion for the year. And we expect noncontrolling interest, or NCI, to net to approximately $250 million in each of Q3 and Q4 of this year and be approximately $1.1 billion for '27 and '28 on a GAAP basis. Turning to CapEx. Due to strong customer demand signals, we're raising our outlook for 2026 and now expect our CapEx to be more than $20 billion, which is up significantly versus our expectations entering the year. We're also aggressively locking in tool purchase orders from our vendors, accelerating our clean room build-outs and actively securing supply of substrates and memory. As a result, we're forecasting 2027 capital expenditures to be significantly above the 2026 levels with the vast majority spent across our U.S. network. In fact, as we look back from 2021 through 2026, our total capital spending in tools and space in the U.S. is approaching $100 billion, significantly higher than any other semiconductor company over that time frame. We remain committed to tightly matching our expenditures with customer demand and remain financially disciplined as we capture the growth ahead. In closing, Q2 was another strong quarter financially and operationally. The client TAM is unfolding as expected, and server CPU demand continues to far outpace available supply. Emerging markets in physical AI, purpose-built silicon, advanced packaging and external wafers are each multibillion-dollar annual revenue opportunities for us in the not-too-distant future. I'm confident in our ability to leverage our broad IP portfolio to solve our customers' most pressing needs and drive long-term value for our shareholders. With that, I'll turn it over to John to start the Q&A.
Thank you, Dave. With that, Jonathan, can we take the first question?
分析師問答
Certainly. And our first question for today comes from the line of Ben Reitzes from Melius Research.
I wanted to ask about the CapEx increase by about $3 billion this year and significantly next. What does that imply for your foundry customers? Does it mean that you've received some part orders for 14A or 18A-P? And then what does it mean for packaging, if you could delineate between the two?
Okay. Let me just write that down. Okay. So let me take the second one first. The CapEx is fairly broad-based. It's going to include advanced packaging. As Lip-Bu talked about, we're pretty excited about our prospects on EMIB-T. And so we will be investing in that. That said, the cost of a fab for the front end is much more expensive than a packaging facility. So it will be skewed towards the front end. But nevertheless, both of them will be important to us. As it relates to customers, I would just say that this increased investment is a signal of our confidence in customers across all of our business units. We feel very confident, particularly in places where we've gotten long-term agreements that we now have the signal to be able to forecast out what the outlook looks like for the next few years in terms of demand, and we're putting forth the capacity in anticipation of that across all of our businesses. That said, as I said in the prepared remarks, we remain very disciplined around spending and Lip-Bu, I think, has beaten that into us that we will put CapEx in place when we feel very confident we can generate a very good return on it. These fabs that we're building, obviously, initially from a cash flow perspective are net cash out. That's why you see the CapEx going up next year. But over time, they generate significant return. And particularly now as we migrate towards a model where we keep these processes on longer, the returns are quite significant. In fact, we're seeing that in Intel 10 and 7 today given the length of time we've had those processes in place.
Ben, do you have a quick follow-up?
Yes. Your competitor today talked about raising the CPU TAM by 2030 to $220 billion with a CAGR, I believe, of 45% or something really great. Can you comment on that? Are you seeing the same thing? And do you think — what do you think of either that number or the growth rate and your ability to fulfill it?
Yes. I mean without putting a number out there, I mean, we obviously recognize that this is a strong market. It's going to grow significantly. Lip-Bu has talked in the past about the ratio of CPU to GPU going up. And we now believe we're almost in parity at this point and could eventually even skew more to CPUs on a unit basis. So I think this market is great. We think we have a strong position in the market and opportunity to capture a significant share of that market. It's going to be a big number, whether anybody can actually predict exactly the number and pinpoint the number, I'm not sure they can. But from all the inputs we're getting from our customers in terms of the level of spend and also the long-term agreements we put in place and the visibility we've gotten, we feel like the growth is going to be significant.
And our next question comes from the line of Joe Moore from Morgan Stanley.
I guess following up on the CPU comment. What do you think happens to your market share in the server space? You have fabs now, which would seem to be an advantage. Do you think that helps to gain share this year? And then as you think about the next five years, your competition versus both AMD and also versus ARM, just how are you guys thinking about the prospects for regaining lost share?
Yes. Maybe I can start first, and then Dave can chip in. So I think clearly, it's a good question. I think I mentioned earlier, the demand is quite strong. And then in terms of agentic AI and inference, the parity ratio of CPU to GPU is changing. And so I think the demand is strong. And I think the challenge right now is more how to grow our supply and to meet the customer requirement. But having said that, clearly, on the server side, data center side, I think we have a strong roadmap. We have Clearwater Forest, Diamond Rapid and also Coral Rapid with the SMT involved. So I think we continue working on improving competitiveness against our competitors. And I think clearly, it's very important to drive improvement in single-thread and also multithreading. Multithreading will be coming in Coral Rapid. So I think across the board we have been really driving that. I think regarding your second question in terms of ARM, clearly ARM is a great partner for us, and we have a strong relationship there. Rene and Mas are good friends of mine. We really focus on not just the ARM-based CPU and also on the ASIC foundry side; they can be a great partner and customer, especially in the IP front. So all in all, I think we compete quite well. We have a strong product roadmap. In some areas we are still behind, but we are catching up very fast, and we try to leapfrog some of the CPU architecture. We are putting major effort into it, and time will tell.
Joe, do you have a quick follow-up question?
Just separately, on the CapEx, is there still a sort of net versus gross aspect to that? Or is that — are you kind of more clear on that? And then how do you think about how you're apportioning that between internal and foundry? Is there a point where you'll be able to delineate that for us? How much is for external customers?
Right. Yes, there is a gross-to-net. At this point now, most of the gross-to-net is AMIC or the investment tax credit. It's running in the low single-digit billions right now, although as we progress, I think we'll start to see that become bigger in terms of chunkiness. At the end of the day, it's a timing thing. But at the end of the day, we're getting roughly $0.35 on the dollar back from the investment tax credit on everything we invest in the U.S. And as I said in my prepared remarks, the lion's share of our capital spend is devoted to the U.S. So it should be pretty significant. But like I said, it will be a delay factor because we've got to build the factory and then we can start claiming tax credits on the factory. And then we have to get it — once we're putting the tools in, it's got to be all the way to production ready before we can claim any sort of credits on the tools. And so — and then you, of course, have to file it with the IRS. So there's a bit of a delay from the time you spend to the time you get it back. But yes, there is a gross-to-net. I think from an investment perspective, I'm not sure we look at it exactly that way, and we really look at how many wafer starts we want in a given node, and we invest to the wafer starts we're looking for on the front end. We have a point of view based on all the demand drivers within the business of what that number looks like, and that's what we're putting in the purchase requisitions with our suppliers to be able to meet those expectations. And of course, we remain nimble as we get more information and we progress. On the packaging side, we are already getting significant backlog. So we already know we need to ramp that up. And that's partly our own internal manufacturing facilities, but we also need to procure substrates from vendors, and there's some requirements there in terms of putting money up in advance of getting the substrates. So we're making those investments probably more quickly right now to get ourselves ready for that.
Jonathan, we have next question please.
Our next question comes from the line of Stacy Rasgon from Bernstein Research.
I wanted to ask about clients. I think everybody kind of had an inkling that data center was going to be pretty strong. I was surprised at the client strength. And I know you talked about demand holding up pretty well. I know you're ramping Panther on some of the new products. But I guess, was it all pricing that drove that revenue strength? Or was there something else going on? And I guess if you could maybe comment on what your expectations are for client in the second half relative to the end market that sounds like it's going to be worse than typical as we move into Q3 and Q4.
Yes. I mean client obviously exceeded expectations. I would say it was largely ASP, of which some of that was mix related. Some of that was our own like-for-like changes in ASPs where we thought we had seen some inflation on our cost and needed to pass that on to the end customer. When you look at it year-over-year, it's definitely down versus the prior market that had a very strong 2025 because of the Windows refresh. So it was coming off a little bit as it relates to that. And then, of course, memory in terms of cost and even availability, I think, has caused that market to be a little softer. But we've kind of skewed the mix to the higher end, and that helped a lot in terms of ASPs and helped it perform. I'd say as we look into next quarter, probably on a revenue basis, it's going to be kind of flattish for us. Overall, CCPG is likely to be up a bit, but I think it will be driven by good growth on the edge and client will be flattish. Underlying that, the market is softer. I think that market is struggling with the memory dynamics in the marketplace. So we think it will be down in the quarter, but because they have been living relatively tight with inventory on CPUs, we likely would see some build in CPUs for Q3. And then Q4, I think it will be — we'll start to experience that in our own business and be soft. The good news for us is we need it because we need the CPUs for the data center side. We can't fulfill the demand, as Lip-Bu was mentioning. So we'll pivot as much of the production as possible over to CPUs and data centers to try to do our best to catch up to what is a pretty significant difference between our ability to supply and the demand out there.
Stacy, do you have a quick follow-up?
I do. You talked about the inventory charges in client. I guess what were those and how big were they? And if I look at that, you're kind of guiding gross margins sort of flattish in the next quarter. If I take out the charges, does that imply that the gross margin guide is actually down sequentially?
Okay. So let me unpack it. So just — sorry, Stacy, you're talking about inventory on our balance sheet for client? Are you talking about...
You talked about inventory charges that the margins in the client itself even on the revenue...
Sorry, I didn't follow that. Yes. Okay. Yes, we took — we had some products where they weren't fully completed from a match set perspective. It just made better economic sense for us to try to pivot more to some of our other products and not complete them given the challenges around match set. So given it was somewhat stranded inventory, we wrote that down. And you're right, we are guiding flat quarter-to-quarter. So obviously, we get a lift from not expecting to have that write-down in the second quarter. The offset of that is even though Panther Lake and Granite are doing better in terms of their cost quarter-to-quarter, they're also becoming a really significant part of the mix, and they're still below the corporate average because they're still relatively early in their life cycle. So that's weighing the margins down a little bit offsetting the lift we get in terms of the reserves, and that's why we think things will be flat. But eventually, both of those things turn into a tailwind. As yields improve on 18A further, we do expect margins to improve on Panther Lake, and they will be above the corporate average, and that will help start lifting the margins there. And obviously, we're keenly focused on improving gross margins over time. I would say, while there's a lot of puts and takes in gross margins, our number one goal this year, which everyone in finance here can attest to because I was a dog with a bone on this, was to get gross margins comfortably into the 40s in every quarter. And I think the team did a really good job getting there, at least for the first two quarters, and our outlook for Q3 would suggest the same. So our goal is to be solidly in that, and then we can kind of pivot from there and look to improve the gross margins off of that base.
Jonathan, we have the next question please.
Our next question comes from the line of Timothy Arcuri from UBS.
Dave, you made a comment that there's a lot of capacity coming online toward the end of this quarter. So I guess that implies a pretty big step-up in revenue for Q4. If I assume that you're still kind of undershipping the market. I think you implied you're undershipping at least last quarter by like more than $1 billion. So if in September, you're undershipping by a similar amount, then it would seem to suggest that Q4 is going to be up pretty big. Is that the right way to think about it? Can you just provide some puts and takes around that?
Yes. I mean, obviously, we only guide one quarter out, which would be my standard response. That said, yes, I mean, of course, if we are able to start to see inventory improve or supply improve towards the end of the third quarter and into the fourth quarter, we would expect a lift from that. I would point out that while things will improve, we will not catch up. We will be behind in the fourth quarter. And I think the team internally is tight with wafers internally. I'm not — I wouldn't dismiss that, but they have done a very good job trying to meet demand and upside us, upside Lip-Bu and I a bit in terms of what they've been able to accomplish. But our supply is a combination of the wafers that we manufacture internally plus advanced packaging like substrates. We've got T-glass. We have memory. And so procuring all of that is also a bit of a choke point for us. In fact, I'd say some of those areas are probably our most challenged parts of our supply chain. So we're working to improve that. While front-end wafers can be a little bit more linear in terms of our improvement, some of those are a bit chunkier. And so we start to see some of the logjam break towards the end of the third quarter, which is why we're more flattish this quarter but see upside in the fourth quarter.
Tim, do you have a follow-up question?
I do. Yes, Dave. So the year-over-year drop-through on gross margin was pretty good in March and June. Based on the guidance, it kind of falls back to the low 50s, which is sort of within your 40% to 60% that you've talked about for drop-through. Is that still the right way to think about it? I mean I'm just trying to see if you can give us some puts and takes into next year.
Yes. I mean longer term, I think we'd expect that to be some of the fall-through somewhere in the 40% to 60% range. Every quarter has unique dynamics to them that affect whether you're at the low end of that range or at the high end of the range or at the midpoint. But I think it's a relatively good rule of thumb for us.
Jonathan, can we have the next question.
Certainly, our next question comes from the line of Vivek Arya from Bank of America Securities.
Lip-Bu, you mentioned increased confidence in engaging with external foundry customers. I'm curious, when will that confidence be backed by actual customer announcements? And then related to that, I think you are planning to raise CapEx. How much should we be expecting a CapEx increase for next year? And how much of that CapEx increase is for external customers versus just expanding capacity for your internal needs?
Yes, thank you so much for the question. I think I will address the confidence that I have, and then I think Dave will talk about the CapEx increase. So first of all, for 18A, 18A-P is in risk production now and will be ready before the end of this year. Clearly, we see about a 5% enhanced performance versus 18A. And 18A yield and production, I think we're starting to see strong metrics. For Panther Lake, we can see that in the ramping in the volume. But let's focus on 14A. For 14A, our PDK 0.5 is complete and PDK 0.9 for 14A is on track for October. That's a very important milestone. I see the yield for 256-bit SRAM and the defect density and performance ahead of the schedule that I set for my team. They met those targets and I'm seeing that. Risk production for 14A will be in the second half of 2027 and then commit to volume production in 2028. So I think all in all, the engagement with the customer and the feedback have been very positive. The moment they start to see the 0.9 PDK and the yield, they're starting to get excited about what kind of product they want to run and how much capacity we can provide them. Those are very positive signs that they are serious about going forward. That's why, as I mentioned earlier, I don't put CapEx in place unless I see yield performance, the IP is ready to serve the customer and customer engagement is at the expected level. Then Dave and I start to put our CapEx to work.
Yes. Maybe it might make sense to start with 2026 and just unpack CapEx so it's clear. We invested a lot in space over the last few years. So we are in a very good place in terms of space. There is still obviously a little investment to facilitize some of the factories, but it's relatively modest. So most of the CapEx dollars at this point are going to tooling. We'll increase tooling in 2026 by 40% relative to 2025. So we're investing a significant amount in tooling. It's where you might expect: Intel 3, 18A, 18A-P. I purposely didn't mention the number for 2027 because we're still working out the exact details of what number we land at. Typically, in this industry, you release that number at the very beginning of the year. But I did want to give investors at least a line of sight to expect that the number will be up. You'll just have to bear with me; I think we need another quarter or two to really solidify the number. And then where it's going, as I said, it's going to both. We're investing for all of our business units, both internally and externally. We take a holistic view of what our wafer demand will look like from all of those customers, and we build capacity that is aligned with that view.
Vivek, do you have a quick follow-up?
Yes. So the follow-up question, Dave, is on the balance sheet. As you are planning these investments for the back half of the year and into next year, how are you thinking about the balance sheet? Do you think that the success you are seeing in your product business on the CPU side is enough to kind of fund a lot of these investments? Or will something else be required?
Yes, it's a good question, Vivek. We feel like we're in a really good place from a balance sheet perspective. We have over $30 billion of cash and a $10 billion revolver. So we've got $40 billion of liquidity. That enabled us to delever, which we felt was important to keep us solidly in investment-grade territory. Obviously, the fact that revenue and profitability and EBITDA are all expanding helps a lot in terms of the cash flow that throws off to the business. Additionally, we have roughly $10 billion of what I call noncore assets that can still be monetized on the balance sheet. Although we're not anxious to do anything there, that's available to us in the event that we need it. We have seen our customers willing to invest with us, and we've had prepays from customers that have enabled us to unlock capacity. That said, if we're super successful, which we're driving to, we may need to tap the capital markets to drive some more investment. If we need to do that, we'll certainly keep the shareholders apprised.
Jonathan, we have the next question please.
Certainly. Our next question comes from the line of C.J. Muse from Cantor Fitzgerald.
I guess maybe a follow-up to a prior question. And Dave, I know you don't want to guide out more than a quarter, but curious how we should be thinking about shape of the server kind of revenue recovery here as you bring on capacity both this year and next? And as part of that, how should we be thinking both from a unit and an ASP perspective as we go through the second half of 2026 and into 2027?
Yes. I think step back and say all of our server wafers are procured internally for the most part. Some ASIC stuff isn't, but the most important node for us on servers is Intel 3 because that's how we produce Granite Rapids. We're seeing tremendous demand there. Granite Rapids is extremely tight because the reception has been fantastic. We are building capacity there over time. It will be a little bit chunky, but I think in general, we have a pretty good ramp of Intel 3 planned for the rest of this year and next year. The challenge is it's not just the front end that we have to expand capacity; we also have to expand capacity on the back end. That then gets into areas that are tight like substrates. We're working to expand that. I think the team did a good job in the first half of the year getting more capacity, but more work is needed to drive that to the levels we would need. From a market perspective, on a units basis, we think the growth rate looks quite good. Units are getting more weighted average core counts, and generally, this market is priced on an ASP per core basis. As you get more cores, you easily get more ASP uplift. That will be a component of revenue growth in this business and gives us confidence that this will be well north of a double-digit CAGR for us over the next few years.
C.J., do you have a follow-up question?
I do, John. Going back to CapEx, and I know you don't want to guide today for next year, but I'm curious if there's a framework in your mind as you balance meeting customer needs as well as your desired free cash flow goals. Is there anything we should interpret around that to help us as we build our model to think what the appropriate CapEx is within that? Or if you've got the signed contracts with customers, are you just going to build it?
We're going to be thoughtful around it. If you look at the core business and what we think we can do from cash flow from operations, even with stepped-up CapEx and the offsets that we get from AMIC, our cash flow actually looks pretty good. But we are likely to have to make investments in the back end as well, particularly with third parties. That could be a drag to cash flow next year and make getting it to a positive number a little more challenged. That said, all of the investment we're making has tremendous ROI. As long as we feel confident in the growth rate, in pricing and cost structure for those products, and we know that when we make investments in these nodes, the lifetime of these nodes is significant and almost always drives a good ROIC, we'll make the investment. We're just going to be very careful about making bets ahead of customer commitments. I think that's the most significant change with Lip-Bu: until we really know that we've got the customers, we don't want to put a significant amount of capital. Given our confidence around next year, we must have significant confidence in our customers that we wouldn't be putting the POs in place today if we didn't.
Jonathan. We've got time for one last question.
Certainly. Then our final question for today comes from the line of Aaron Rakers from Wells Fargo.
I guess the first question is on the ASIC business. Based on what was disclosed last quarter, it's about a $1.2 billion run rate business now growing well for the company. How do we think about the diversity of that business? You've announced Fortinet. I'm curious how you're thinking about the growth profile of that business? And maybe also the margin profile of ASICs as they expand.
Let me start first. This is a massive opportunity. Potentially it's over a $100 billion TAM market. We have a unique opportunity to offer, besides our advanced design capability using our CPU and xPU, a strong IP portfolio and advanced packaging that are very unique in terms of vertical integration and necessary layers. Combined, they can really drive leadership; many new AI technologies need this packaging technology plus our advanced silicon process technology. That combination gives a lot of opportunity for many purpose-built silicon products that companies need. One example is our recent announcement with Fortinet for security ASICs. We will drive next-generation security processors with higher performance. And, of course, we all know about the Intel IPU providing to some hyperscalers, and that is a tremendous opportunity. I mentioned it's grown roughly 3x year-over-year in the quarter, so it's a great opportunity for us. Dave, do you want to add?
What was the question? Growth rates? I would say that today, we're probably running at about a $2 billion run rate or at least approaching a $2 billion run rate for that business. We think in the not-too-distant future, we'll be at a $4 billion run rate for that business. So I think that gives you a sense. We think the TAM is large, and given our IP portfolio and what we bring to bear, we deserve a good chunk of that. Stay tuned.
Aaron, do you have a quick follow-up question?
Yes. There's a lot going on in memory and changes to memory hierarchy and architecture. There's been some recent news about Intel's own development work, I think a technology called Z-Angle Memory, cross-batch memory. As you strategically look at the role memory plays in these architectures and scaling compute, does Intel serve a bigger role in that? Are there internal developments and even opportunities in memory over time?
Good question. A couple of things. First, memory becomes a big supply constraint challenge. We're collaborating with the three big memory vendors. That's very important to serve our customers as our number one priority. Second, Intel has a rich history in memory. Recently, we hired Seok-Hee Lee to join us; he used to be CEO of SK Hynix. Memory becomes the bottleneck for a lot of AI infrastructure and a pain point for customers. We're also looking at how other areas we can integrate compute and memory and how stacking and utilization of memory can be more efficient. There are many areas we are working on. Stay tuned; we will keep you posted. With that, I think my closing remarks: I want to thank everyone for joining us today. We made good progress this quarter on our journey to transform Intel to the new Intel, but we still have a lot to do ahead of us and look forward to seeing many of you throughout the quarter and providing additional updates in October.
Thank you, ladies and gentlemen, for your participation in today's conference. This does conclude the program. You may now disconnect. Good day.