Intel（Minutes）: Achieve breakeven for foundry services by the end of 2027 (24Q4 conference call)

Intel released its Q4 2024 financial report (ending December 2024) after the U.S. stock market close on January 31, 2025, Beijing time. The key points from the conference call are as follows:

Below is the summary of Intel's Q4 2024 financial report conference call. For the report interpretation, please refer to “Intel: Layoffs and Cost Reductions Show Results, But Growth Remains a Challenge?”

1. $Intel(INTC.US) Core Information Review:

2. Detailed Content of Intel's Financial Report Conference Call

2.1 Executive Statements on Core Information:

1. Financial Performance Overview

Q4:

Revenue: $14.26 billion, a year-on-year increase of 3.14%, exceeding the high end of guidance.

Earnings Per Share (EPS): $0.13, a year-on-year increase of 8.33%, exceeding the guidance expectation of $0.12.

Gross Margin: Non-GAAP gross margin of 42.1%, exceeding guidance expectations by 260 basis points.

Operating Cash Flow: $3.2 billion, a quarter-on-quarter decrease of approximately $900 million, mainly due to cash expenditures for restructuring in Q3.

Capital Expenditure: $6.3 billion, adjusted free cash flow of negative $1.5 billion.

Full Year (FY 2024):

Revenue: $53.1 billion, a year-on-year decrease of 2.1%.

Gross Margin: 36%, a year-on-year decrease of 760 basis points.

Earnings Per Share (EPS): -$0.13, a year-on-year decrease of $1.18.

Operating Cash Flow: $8.3 billion, capital expenditure of $24 billion, adjusted free cash flow of negative $2.2 billion.

2. Performance of Each Business Segment

Intel Products:

Q4 Revenue: $13 billion, a quarter-on-quarter increase of 7%.

Client Computing (CCG): Revenue increased by 9% quarter-on-quarter, with a significant slowdown in customer inventory digestion.

Data Center Infrastructure (DCI): Revenue slightly increased quarter-on-quarter, with stable demand for traditional servers.

Network and Edge (NEX): Revenue increased by 7.5% quarter-on-quarter, growing over 20% from the Q2 low.

Operating Profit for Intel Products: $3.6 billion, accounting for 28% of revenue, a quarter-on-quarter increase of $300 million.

Intel Foundry Services: Q4 Revenue: $4.5 billion, a quarter-on-quarter increase of 3%, mainly due to an increase in EUV wafer proportion and growth in IMS equipment sales EUV wafer revenue: Increased from 1% in 2023 to over 5% in 2024.

Intel's foundry division operating loss: $2.3 billion, significantly improved quarter-over-quarter, impacted by a $3.1 billion impairment in the third quarter.

Other businesses:

Mobileye: Fourth-quarter revenue of $490 million, a 1% increase quarter-over-quarter, with an operating profit of $103 million.

Altera: Fourth-quarter revenue of $429 million, a 4% increase quarter-over-quarter, with an operating profit margin of 21%.

3. Future Outlook

First quarter of 2025 (Q1):

Revenue expectation: $11.7 billion to $12.7 billion, a quarter-over-quarter decline of 11% to 18%.

Gross margin (non-GAAP): Expected to be around 36%.

Tax rate: 12%.

Earnings per share (EPS): Expected to be $0, on a non-GAAP basis.

Full year (FY 2025):

Gross margin: Expected to gradually improve from the low point in the first quarter.

Intel product gross margin: 51% in 2024, expected to decline in 2025 due to changes in product mix.

Intel foundry services gross margin: Expected to improve with the increase in EUV wafer proportion and advanced packaging growth.

Operating expenses (OpEx): Target of $17.5 billion, with further reductions in 2026.

Non-controlling interests (NCI): Expected to be approximately $0 in the first quarter, with an annual expectation of $500 million to $700 million.

Capital expenditure: Expected to grow capital investments to approximately $20 billion for the year, with net capital expenditures of $8 billion to $11 billion.

4. Strategic and Operational Progress

Intel Foundry Services:

Technical progress: The 18A process has made good progress, with Panther Lake products set to launch in the second half of 2025, supporting increased production and profitability in 2026.

Customer trust: Gradually gaining customer trust by providing competitive technology and reliable execution capabilities.

Financial goal: Aim to achieve operational breakeven for Intel Foundry Services by the end of 2027, ultimately achieving a good return on invested capital (ROIC).

Intel Products:

Client business: The Lunar Lake product has been launched, with strong market competitiveness despite higher costs. The Panther Lake, expected to launch in the second half of 2025, is anticipated to significantly enhance performance and profit margins.

Data center business: The Granite Rapids product has achieved initial success, with Clearwater Forest planned for launch in the first half of 2026, aiming to stabilize market share and enhance competitiveness.

AI data center: Simplifying the product roadmap and focusing resources on developing Jaguar Shores to provide a complete rack-level solution.

2.2 Q&A

Q: Regarding Intel's Data Center business (DCAI), can you elaborate on how Granite Rapids is performing in narrowing the gap with competitors? What is the launch plan for Clearwater Forest in 2025? When can we clearly see Intel narrowing the gap with competitors in the data center business?

A Granite Rapids is a good start for Intel's data center business, and it has indeed narrowed the gap with competitors. Customers are excited about it, and we are beginning to see the product's competitiveness in terms of shipments. However, we clearly recognize that continuous improvement is still needed. Subsequent products like Diamond Rapids will further enhance our competitiveness.

Regarding Clearwater Forest, we see it as a niche in the data center market, primarily targeting specific needs. The product is scheduled to launch in the first half of 2026, and the 18A process has performed well in terms of performance and yield, but due to its complex packaging requirements, the launch is set for 2026. We expect Clearwater Forest to be an excellent product and continue to narrow the gap with competitors. However, this is an ongoing process, not a one-time event.

Q What are the main reasons for the quarter-over-quarter decline in gross margin in the first quarter? How is the trend of gross margin expected to be for the whole year?

A The main reason for the quarter-over-quarter decline in gross margin in the first quarter is the decrease in revenue, which is expected to drop by about $2 billion. Additionally, there were some one-time factors in the fourth quarter, such as revenue exceeding expectations and the signing of the CHIPS agreement, which allowed us to offset some subsidies as costs, thereby boosting the gross margin in the fourth quarter. However, these factors do not exist in the first quarter.

For the entire year of 2025, the gross margin of Intel Products will face pressure, especially since the Lunar Lake products have higher costs, which will lower the gross margin. Meanwhile, the gross margin of Intel Foundry will improve as the proportion of EUV wafers increases and advanced packaging grows. It is expected that the gross margin for the whole year will gradually improve from the low point in the first quarter.

Q All three product segments are expected to see a quarter-over-quarter decline in revenue in the first quarter, although the PC inventory digestion and tariff factors mainly affect the client business. Why do the data center and network edge (NEX) businesses also experience a similar decline?

A We are cautious about all markets, as macroeconomic uncertainty affects all businesses. Additionally, seasonal factors also impacted most businesses in the first quarter. Therefore, the combination of macroeconomic uncertainty and seasonal factors led to the revenue decline in the data center and network edge businesses.

Q Will intensified competition have a lasting impact on the gross margin for the whole year? How to cope with pricing pressure in the client and data center businesses?

A Intensified competition is a reality, especially in the client computing (CCG) sector, where new entrants have brought more competition. Although the Lunar Lake products perform well, the higher costs put some pressure on the gross margin. However, we are committed to stopping the loss of market share in the client market and striving to win every bit of market share. The same goes for the data center business; Granite Rapids is a positive start, and we need to stop the loss of market share in the data center business. We need to be proactive, win market share, and show customers that they can succeed with us Q Has the strategy of Intel Foundry Services (IFS) been adjusted under the new joint leadership?

A We are no longer investing excessive capital upfront; instead, we are deploying capital more cautiously to ensure we are more prudent when making commitments to customers. We aim to exceed expectations at every level, including investors, customers, and suppliers. Our primary goal remains to build world-class foundry services, and we believe there is a need for another player in the leading semiconductor manufacturing field, especially in the United States, which aligns with the interests of the U.S. government. We will continue to focus on this, but we will be more cautious in capital investment and customer collaboration to achieve the best return on invested capital (ROIC).

Q Given the weakness and uncertainty in the first quarter, what is the path to achieving positive free cash flow growth in 2025?

A We have not provided specific guidance for the full year of 2025, but we achieved good operating cash flow in 2024 through strong working capital improvements and corresponding capital expenditure adjustments in challenging circumstances. Although 2024 is expected to be negative, it is closer to zero than anticipated. In 2025, we will continue to focus on operating cash flow, effectively manage working capital, and adjust capital expenditures based on new outlooks. We also anticipate over $10 billion in offsetting factors that will help us improve free cash flow. While it is not convenient to provide an adjusted free cash flow forecast for the full year at this time, this is a key focus for us, and we hope to assist our deleveraging through the monetization of non-core businesses, which will also be a focus in 2025.

Q You mentioned an adjustment in expectations for Falcon Shores in your prepared remarks; what is the reason behind this? What needs to be done to become competitive in this area?

A This conclusion is based on in-depth collaboration with the team over the past six weeks, a review of the product roadmap, and an assessment from competitive and execution perspectives. Extensive communication with customers has also made me aware of their views on competitiveness and their expectations for the product. We have learned a lot from the Gaudi product, but the key is that merely providing chips is not enough. We need to offer complete rack-level solutions, which is what Jaguar Shores can achieve. Falcon Shores will play a role in developing system, network, memory, and other component functionalities, but what customers truly need is a complete rack-level solution.

Additionally, discussions around DeepSeek this week have made us realize that no single solution can meet all needs. Therefore, I am reviewing our product roadmap and considering how to leverage the vast amount of existing IP and assets from Intel's product divisions to address this market. We have excellent CPUs, GPUs, ASICs, FPGAs, etc., and we need to figure out how to integrate these resources because when customers face constraints, they will look for different ways to deploy technology. This is also a huge opportunity, and I am exploring whether we can achieve breakthroughs in this area.

Q You mentioned that tariffs might prompt some customers to place orders early; how common is this situation? Is it based on conservative considerations or actual evidence? A We have a fairly accurate understanding of our customers' quarterly demand. In some cases, customer order volumes have exceeded our expected absorption capacity. This has made us realize that they may be doing this for some reason, and we know that tariffs are a focal point for many customers. This situation mainly occurs in the Asia region, and we believe customers may be engaging in some hedging operations to bring forward fourth-quarter revenue while distancing themselves from the first quarter. However, it is currently difficult to extrapolate this trend beyond this quarter due to the many uncertainties surrounding the tariff plans.

Q Regarding gross margin, you mentioned that the incremental gross margin is about 60%, but this is based on the fourth-quarter guidance of 39.5%, while the actual fourth-quarter gross margin was 42.1%, and now it has dropped to 36% in the first quarter. So, how should we consider future incremental gross margins from here?

A At this stage, our usual incremental gross margin rule is 60%. However, there is a dynamic factor in 2025, which is the gross margin pressure from products like Lunar Lake, which may lower the range to a more reasonable incremental gross margin of about 40% to 60%. Entering 2026, with more 18A production from Panther Lake, we expect the incremental gross margin to exceed 60%.

Q You mentioned that capital expenditures have been adjusted to $20 billion, but there is a $1.2 billion outflow in the financing section of the cash flow statement. I want to clarify whether capital expenditures are really decreasing? Will the items in this financing section continue to expand?

A The capital expenditure forecast for 2025 is $20 billion, primarily because we are better utilizing our assets under construction. Our philosophy has always been to invest early, and we have accumulated over $50 billion in assets under construction that have not yet been deployed. Therefore, we are pushing the team to absorb as many of these assets as possible and limit external procurement. This will allow us to reduce capital expenditures to around $20 billion. We are not engaging in any complex financing operations, but for transparency, capital expenditures include two aspects: one is the placement of equipment orders, and the other is cash payments. Therefore, we are indeed working to improve capital expenditures by negotiating payment terms with suppliers to reduce our capital expenditures. But to be honest, when we actually deploy assets and start depreciation, we have already spent the funds in all cases because that money goes into assets under construction and may stay there for about 9 months until it is actually put into use.

Q In the data center server CPU market, is the change in market share for Intel compared to x86 competitors due to better design or better manufacturing processes? Would outsourcing more to external foundries help regain market share? Have we seen some market share shifts in the enterprise market?

A From the perspective of data centers and competitiveness, Granite Rapids has made a good first step in narrowing the gap with competitors, but we still have a gap to close. Therefore, we need to focus on delivering Diamond Rapids. The early feedback on these products has been very positive. When it comes to external manufacturing, I have always been very transparent about my views I believe you must have the right products, the right processes, and deliver within the right time window. Currently, about 30% of Intel's manufacturing is done through various external partners. This may be our highest level at the moment, but it will never be zero. What I can tell you is that the 18A process is progressing well, and it has won the business for Panther Lake and Clearwater Forest. However, when I consider the future development of the data center business and the future roadmap, I ask myself this question every time. Therefore, I do not think it is unimaginable to outsource data center products. If it means I can provide the right products and performance to customers within the right time window, then I might consider doing so.

Q Regarding non-controlling interests (NCI), you mentioned $500 million to $700 million for 2025, then growing to $1.2 billion to $1.4 billion in 2026. Will this growth continue? How can it be reasonably modeled, as the less outsourcing there is, the larger the share of in-house manufacturing, and the more reversals of the NCI portion will occur, which is a headwind for reported earnings per share (EPS), right?

A That's a good question. First, NCI is influenced not only by Skip but also by Mobileye's revenue, and the NCI will be further exacerbated after the sale of stakes in companies like Mobileye and Altera. Therefore, there are many factors that affect NCI, making predictions somewhat difficult, as you need to know the exact shares of each asset and the production situation of each factory.

Our guidance for 2025 and 2026 is based on our current understanding, and there may be further increases in 2027, but it is currently difficult to determine exact figures.

Q In your prepared remarks, you mentioned looking forward to working with the Trump administration. Can you provide more details? Have they reached out proactively? Who is leading these discussions on Intel's side? Can you share some insights about the discussion topics and what they are particularly interested in?

A We have maintained good communication with them, and since the election, we have been in contact with their team at various levels, including at the CEO level, and we have a strong government affairs team that communicates with them daily. I am very optimistic about their outlook on bringing semiconductor manufacturing back to the U.S., which is a very positive signal for us.

To be honest, we have never left the U.S. Therefore, we are in a favorable position in this regard. They recognize the value of advanced semiconductor manufacturing R&D in the U.S., which is also positive for us. They want to see more jobs return to the U.S., and we provide high-paying, high-tech jobs, which clearly aligns with their interests. But more importantly, this is about security, whether it is supply chain security or reliable manufacturing for the Department of Defense, we clearly have the capability to provide these services for them. As we grow, I expect we will collaborate more deeply with them to achieve their goals.

Q Regarding gross margin, can you be more specific, considering our competitive pricing in server CPUs and client CPUs, as well as the possibility of outsourcing more business to TSMC and others, how is the trend of gross margin expected to be from the first quarter? **

A We tend to maintain competitiveness in the product area. Additionally, as I mentioned, due to the cost structure of Lunar Lake, the gross margin will face certain pressure. Therefore, this will definitely affect the gross margin of Intel Products. For the full year of 2025, there won't be a significant improvement in the gross margin of this business unit until Panther Lake is launched, at which point I believe we will see a better cost structure and have a competitive product that may allow us to relieve some pressure in a highly competitive market.

So, the foundry business will see improvements. The situation will get better with the return of more wafers, especially after the launch of Panther Lake in 2026. We are improving the cost structure of the foundry services business as part of an overall spending reduction plan. This will also help. Then remember, the 18A wafers produced by Intel have a better cost structure and profit margin compared to the price structure, which will have a positive impact on the foundry services business.

Therefore, from a macro perspective, I think the best approach might be to set the incremental gross margin between 40% and 60%, which could be a reasonable approximation for estimating gross margin based on your forecasted quarterly revenue.

Q In the prepared remarks, it was mentioned that the share of EUV wafers will grow from 1% in 2023 to over 5% in 2024. Can you provide a framework for what a successful definition of the share of EUV wafers will be by the end of 2025? And remind us again, what are the differences in cost structure and profit margin for EUV wafers?

A Regarding the second part, the price of EUV wafers is three times that of Intel's 3 or 4 node costs. Therefore, when you transition from Intel's 3 or 4 nodes to 18A, there will be a significant improvement in gross margin. This is particularly evident when comparing Intel's 18A to non-EUV wafers.

As for the specific number for the share of EUV wafers by the end of 2025, it is difficult to predict exactly. However, with the launch of Panther Lake in 2025 and the development of Meteor Lake on Intel's 3 or 4 nodes, we expect a significant increase in the share of EUV wafers by the end of 2025.

Q Regarding NCI, I just want to confirm again that when we report earnings per share (EPS) on a non-GAAP basis, NCI is indeed included in that EPS number to ensure I am modeling correctly.

A Yes, NCI is included in the non-GAAP number to arrive at our reported fully diluted non-GAAP earnings per share (EPS).

Q Regarding the goal of achieving breakeven for the foundry services (IFS) by 2027, can you discuss what the underlying assumptions are? Do you think breakeven can be achieved solely with internal wafers, or will external customers also be needed? If the latter, how much revenue would need to be generated from external customers to achieve breakeven?

A We are still committed to achieving breakeven by 2027. This is primarily based on internal wafers from Intel Products, particularly EUV wafers, as they have better profit margins With the increase in the proportion of EUV wafers, this will significantly improve profit margins.

But more importantly, the original intention of creating this independent P&L structure is to drive the foundry service business to focus more on efficiency, squeeze more value from the existing footprint, be more sensitive to capital, and ultimately consider return on invested capital (ROIC) in all decision-making. I believe this has already worked. I often hear the team's shift in meetings, where Michelle and I participate, they are completely focused on how to make money in that business. Therefore, I think this is working. As we develop in 2025 and 2026, I expect we will see more efficiency improvements. Therefore, I am optimistic about our ability to achieve breakeven in 2027. Of course, we want to have external customers, so we also consider a certain proportion of external customer revenue in our assumptions for 2027. But if 18A is as attractive as customer feedback suggests, I think we may exceed expectations in the ratio of external to internal customers. Therefore, these factors will drive us to profitability in 2027. Ultimately, we want to push this business to a profitability level consistent with the foundry industry.

Q Regarding the Panther Lake of 18A, I remember you mentioned in the past that you plan to bring back about 70% of the chips for internal production. Is this plan still valid? What is the flexibility of bringing more or fewer chips back for internal production? Just wanted to understand.

A Yes, we have indeed shifted the design of Panther Lake to internal production of 18A. But as I mentioned before, we decide what the right product, the right process, the right time window, and what can make our customers successful based on each product generation. For Panther Lake, 18A is the right choice. Currently, we are very satisfied with the performance and yield of 18A. Therefore, it will remain on 18A. Then, as we look forward to the next generation client product Nova Lake, that product will actually contain both internal and external chips simultaneously. So you will see computing chips both internally and externally. Similarly, this is about optimization to ensure we can succeed in the market, help our customers succeed, and optimize the entire product portfolio because ultimately, if our customers succeed, we will win more Intel foundry service wafers, which will lead to our success. But I will continue to maintain this balance. As I said, we will also examine the data center product portfolio in the same way.

The risk disclosure and statement of this article: Dolphin Investment Research Disclaimer and General Disclosure