Intel Foundry’s second-quarter performance marks a meaningful improvement in execution, but it does not yet settle the question that will define the company’s manufacturing strategy: can Intel become a trusted, large-scale supplier for chip companies beyond Intel itself?
The numbers show progress. Intel Foundry reported $5.8 billion in quarterly revenue, up from $4.4 billion in the comparable period a year earlier. Its operating loss narrowed sharply, from $3.2 billion to $2.1 billion, while its operating margin improved from negative 71.7% to negative 36.2%.
Those figures matter because they indicate that Intel’s expensive manufacturing rebuild is beginning to produce operational results. Better yields, faster cycle times, and increased fab scale are lowering costs across Intel 4, Intel 3, and Intel 18A production.
But the more revealing number is far smaller: only $293 million of Intel Foundry’s quarterly revenue came from external customers. The overwhelming majority of foundry activity remains Intel manufacturing chips for Intel.
That distinction is critical. A company can run factories more efficiently, fill them with its own products, and reduce manufacturing losses without proving that it has established a successful merchant foundry business. Intel is moving in the right direction, but it still needs external design wins that translate into sustained wafer volumes, packaging revenue, and long-term customer trust.
For Windows users, PC builders, enterprise IT teams, and the broader hardware ecosystem, this matters far beyond Intel’s quarterly results. Intel Foundry’s success or failure could influence future processor supply, competition in advanced chip manufacturing, geographic diversification, Windows PC availability, and the economics behind the next generation of AI-capable devices.
Intel’s latest results give the company a stronger operational foundation than it had a year ago. The foundry segment remains unprofitable, but losses are falling as factory utilization rises and process technologies mature.
Manufacturing economics are unforgiving. A leading-edge fab requires extraordinary spending on buildings, clean-room equipment, lithography systems, chemicals, materials, engineering talent, and long-term capacity commitments. Once a fab is built, it needs high production volume and healthy yields to spread those fixed costs across enough usable wafers.
Intel’s recent gains suggest improvement in several areas:
For consumers, process technology may sound distant from everyday computing. In practice, it shapes the laptops, desktops, workstations, servers, and AI systems that run Windows workloads. A more efficient process can enable processors with lower power draw, better battery life, higher sustained performance, and more room for integrated graphics or AI acceleration.
Intel’s ability to manufacture its own processors at improving yields is therefore not merely a financial milestone. It is a prerequisite for delivering competitive products across the Windows PC market.
That is good news in the short term. Internal demand gives Intel a reliable base load while its process nodes mature. It can also create a positive feedback loop: higher utilization reduces costs, better costs support better product margins, and improved margins help fund further development.
The complication is that internal manufacturing volume does not automatically prove external competitiveness.
A merchant foundry earns its reputation by convincing outside companies that its manufacturing technology, design tools, intellectual-property libraries, packaging options, delivery schedules, and customer support are compelling enough to justify entrusting it with major products. Intel’s internal teams have little choice but to use Intel fabs when corporate strategy directs them to do so. External customers have alternatives.
That is why Intel Foundry’s operational recovery should be recognized as a substantial achievement without being mistaken for final proof that Intel has solved its commercial foundry challenge.
Fortinet is not simply evaluating an Intel process in a laboratory setting. The relationship is positioned as a commercial engagement involving design, packaging, and manufacturing capabilities. Fortinet leads the architecture and front-end design work, while Intel contributes back-end design services and production support.
Intel has described itself as the exclusive manufacturing partner for SP6. That makes the deal more significant than a preliminary process evaluation, a test-chip project, or a non-binding technology partnership.
The arrangement also offers a clearer view of how Intel may initially build its external customer base.
Intel’s work with Fortinet appears broader than that model. It combines:
Many companies need specialized chips for networking, cybersecurity, industrial systems, automotive platforms, cloud infrastructure, and edge AI appliances. Yet not all of them have the internal semiconductor engineering resources of Apple, Nvidia, AMD, Qualcomm, or large hyperscale cloud providers.
Intel can potentially serve these customers with a more complete package: processor IP where appropriate, engineering support, manufacturing, packaging, validation expertise, and supply-chain capability. In this model, Intel competes not only with Taiwan Semiconductor Manufacturing Co. but also with custom-silicon specialists such as Broadcom and Marvell.
That could create a valuable niche. Intel does not need to win every high-profile smartphone processor or AI GPU design to build a viable external business. It can win purpose-built chips where customers prioritize performance, security, integration, domestic or geographically diversified manufacturing, and engineering support.
As businesses deploy more encrypted traffic, cloud connectivity, AI-driven threat detection, and distributed branch environments, security hardware needs to process more data without becoming a bottleneck. Custom silicon is one way to improve efficiency compared with relying entirely on general-purpose CPUs.
For enterprise Windows environments, that can have indirect but important consequences. The security infrastructure protecting Windows endpoints, Active Directory environments, hybrid cloud resources, and remote-access deployments increasingly depends on high-throughput, specialized networking hardware.
A stronger Fortinet ASIC roadmap could improve appliance performance and efficiency. Intel’s participation gives the project a potential supply-chain angle as well, which may appeal to organizations that want more manufacturing resilience in critical infrastructure.
Still, the Fortinet win comes with limits. Intel has not disclosed the SP6 production schedule, expected volumes, packaging technology, fab location, financial terms, or date of volume manufacturing. Until SP6 reaches meaningful production, it cannot be treated as a major contributor to Intel Foundry revenue.
It is a promising proof point, not yet a revenue-scale breakthrough.
Intel 4 is a proven production process with a track record inside Intel’s own product portfolio. That reduces technical risk for a customer such as Fortinet and gives Intel an opportunity to demonstrate commercial execution with an external partner.
However, the future of Intel Foundry depends more heavily on whether customers choose its newer nodes, particularly Intel 18A-P and Intel 14A.
The distinction matters because foundry customers do not choose a process node based only on a transistor roadmap slide. They evaluate a broad set of criteria:
Intel has said that 18A-P has entered risk production. That is an important technical milestone, but risk production is not the same as broad commercial volume. It is a stage in which process technology is tested, refined, qualified, and prepared for customer designs.
Intel’s commitment to 14A high-volume manufacturing is a notable signal. Advanced-node development is extremely expensive, and companies do not commit to that scale of investment lightly. Intel’s management has pointed to improving defect density, transistor performance, process-design kits, and intellectual-property progress as reasons for confidence.
Yet the wording around 14A remains revealing. Intel is committing to ramp the node for its own products, while a major named external production customer has not been announced.
That is the gap Intel must close.
A major outside customer choosing Intel 14A for a high-volume product would carry far more weight than a vague statement of customer engagement. It would demonstrate that a company with other manufacturing choices believes Intel can meet strict requirements for confidentiality, process maturity, design enablement, capacity, packaging, and long-term support.
Until then, Intel 14A remains a technically ambitious roadmap with commercial potential rather than a proven external-foundry platform.
A pure-play foundry, by design, does not compete with its customers in major end markets. Its role is to manufacture chips for others. Intel is different. It owns major processor product lines in client computing, servers, graphics, networking, and AI infrastructure.
That creates a natural concern for outside customers: what happens when Intel’s own products need the same scarce factory capacity, engineering resources, packaging lines, substrates, and advanced manufacturing equipment?
During that period, a customer needs assurance that:
External customers need repeated evidence. They need to see Intel honor commitments through difficult periods, not only when factory capacity is plentiful.
The company’s current internal supply constraints make that test more immediate. Intel is expanding Intel 3 capacity for Granite Rapids server processors while increasing 18A output for new client and server products. Those internal ramps help foundry utilization, but they also highlight exactly why external customers may want contractual and operational protections.
A foundry customer does not want to be treated as overflow demand. It wants confidence that it is a strategic partner.
That spending is immense, but it reflects the economic reality of the semiconductor industry. Leading-edge manufacturing requires multibillion-dollar commitments years before a product reaches store shelves or data centers.
Intel has emphasized that it is matching investment to customer demand rather than building large amounts of speculative capacity. That is financially prudent. It also suggests management believes it has enough confidence in future demand to place major equipment orders.
Still, there is a critical question: how much of that demand comes from external foundry customers, and how much comes from Intel Products?
The company has not provided a detailed breakdown. That limits outside visibility into whether foundry investment is being driven primarily by a growing merchant business or by the need to supply Intel’s own processors.
TSMC has an enormous lead in merchant-foundry scale, customer trust, ecosystem depth, and advanced-node volume. It serves a broad collection of the world’s most demanding semiconductor companies and has built decades of credibility as a manufacturing partner.
Intel is not likely to displace that position quickly. Nor does it need to do so for Intel Foundry to succeed.
A customer may not move an entire flagship design away from its incumbent foundry. But it may want:
Intel’s U.S. and European manufacturing footprint may become a commercial advantage where supply-chain resilience and regional manufacturing are strategic priorities.
But optionality is an intermediate stage, not the final goal. Intel’s defining external-foundry milestone will arrive when a major company chooses Intel as the primary production source for a high-value chip rather than simply as a hedge against capacity risk elsewhere.
A healthier Intel manufacturing operation could improve the company’s ability to deliver competitive processors for Windows laptops, desktops, workstations, and enterprise systems. Better yields and improved wafer costs can support higher-volume availability and potentially give Intel more flexibility in product segmentation and pricing.
The effects may extend across several areas:
But Intel Foundry becoming more effective gives the broader Windows hardware ecosystem another potential source of advanced manufacturing capacity. That matters in an era when chip availability can shape everything from laptop launch timing to data-center expansion plans.
The next phase requires more concrete evidence.
The company needs a portfolio of customers across different industries, process nodes, product categories, and production volumes. It also needs those customers to return for future generations of silicon. Repeat business is one of the clearest indicators that a foundry relationship has moved from experimentation to trust.
The Fortinet agreement also represents a legitimate external customer win. It shows that Intel can combine its design, manufacturing, and packaging capabilities into a commercial offering for a company building specialized silicon.
Yet Intel’s core challenge remains unresolved. The company is still primarily a manufacturer for itself, and its external revenue remains modest relative to the scale required to support a global merchant-foundry ambition.
The real measure of success will not be whether Intel can improve utilization in fabs filled by its own processors. It will be whether outside chip companies choose Intel for major products, remain confident that their intellectual property and production priorities are protected, and commit enough volume to turn Intel Foundry into a durable independent business.
Intel has demonstrated that its manufacturing recovery is gaining momentum. The next breakthrough must be commercial: not simply proving that Intel can make better chips for Intel, but proving that other companies want Intel to make their most important chips too.
The numbers show progress. Intel Foundry reported $5.8 billion in quarterly revenue, up from $4.4 billion in the comparable period a year earlier. Its operating loss narrowed sharply, from $3.2 billion to $2.1 billion, while its operating margin improved from negative 71.7% to negative 36.2%.
Those figures matter because they indicate that Intel’s expensive manufacturing rebuild is beginning to produce operational results. Better yields, faster cycle times, and increased fab scale are lowering costs across Intel 4, Intel 3, and Intel 18A production.
But the more revealing number is far smaller: only $293 million of Intel Foundry’s quarterly revenue came from external customers. The overwhelming majority of foundry activity remains Intel manufacturing chips for Intel.
That distinction is critical. A company can run factories more efficiently, fill them with its own products, and reduce manufacturing losses without proving that it has established a successful merchant foundry business. Intel is moving in the right direction, but it still needs external design wins that translate into sustained wafer volumes, packaging revenue, and long-term customer trust.
For Windows users, PC builders, enterprise IT teams, and the broader hardware ecosystem, this matters far beyond Intel’s quarterly results. Intel Foundry’s success or failure could influence future processor supply, competition in advanced chip manufacturing, geographic diversification, Windows PC availability, and the economics behind the next generation of AI-capable devices.
Intel Foundry Is Improving Its Manufacturing Execution
Intel’s latest results give the company a stronger operational foundation than it had a year ago. The foundry segment remains unprofitable, but losses are falling as factory utilization rises and process technologies mature.Manufacturing economics are unforgiving. A leading-edge fab requires extraordinary spending on buildings, clean-room equipment, lithography systems, chemicals, materials, engineering talent, and long-term capacity commitments. Once a fab is built, it needs high production volume and healthy yields to spread those fixed costs across enough usable wafers.
Intel’s recent gains suggest improvement in several areas:
- Higher manufacturing yields, which means a greater share of produced chips meet specifications.
- Shorter cycle times, reducing the period wafers spend moving through complex fabrication stages.
- Improved wafer costs, helped by more efficient use of fabs and process equipment.
- Greater scale across Intel 4, Intel 3, and Intel 18A, allowing manufacturing overhead to be distributed across more output.
- Better utilization of advanced facilities, particularly as demand for Intel’s own client and server processors rises.
For consumers, process technology may sound distant from everyday computing. In practice, it shapes the laptops, desktops, workstations, servers, and AI systems that run Windows workloads. A more efficient process can enable processors with lower power draw, better battery life, higher sustained performance, and more room for integrated graphics or AI acceleration.
Intel’s ability to manufacture its own processors at improving yields is therefore not merely a financial milestone. It is a prerequisite for delivering competitive products across the Windows PC market.
Internal Demand Is Both a Strength and a Complication
Intel’s own product divisions are helping fill the fabs. Stronger demand for client CPUs, Xeon server processors, and purpose-built silicon creates more volume for Intel Foundry and helps the company improve manufacturing utilization.That is good news in the short term. Internal demand gives Intel a reliable base load while its process nodes mature. It can also create a positive feedback loop: higher utilization reduces costs, better costs support better product margins, and improved margins help fund further development.
The complication is that internal manufacturing volume does not automatically prove external competitiveness.
A merchant foundry earns its reputation by convincing outside companies that its manufacturing technology, design tools, intellectual-property libraries, packaging options, delivery schedules, and customer support are compelling enough to justify entrusting it with major products. Intel’s internal teams have little choice but to use Intel fabs when corporate strategy directs them to do so. External customers have alternatives.
That is why Intel Foundry’s operational recovery should be recognized as a substantial achievement without being mistaken for final proof that Intel has solved its commercial foundry challenge.
Fortinet Is a Genuine External Customer Win
The most important external development in Intel’s latest quarter is the collaboration with Fortinet, which plans to use Intel technology for its next-generation Security Processor 6, or SP6.Fortinet is not simply evaluating an Intel process in a laboratory setting. The relationship is positioned as a commercial engagement involving design, packaging, and manufacturing capabilities. Fortinet leads the architecture and front-end design work, while Intel contributes back-end design services and production support.
Intel has described itself as the exclusive manufacturing partner for SP6. That makes the deal more significant than a preliminary process evaluation, a test-chip project, or a non-binding technology partnership.
The arrangement also offers a clearer view of how Intel may initially build its external customer base.
A Hybrid Foundry and Custom-Silicon Model
Traditional merchant foundry business is straightforward in concept, even if it is technically complex in execution. A customer designs a chip, selects a process node, uses the foundry’s design rules and intellectual-property ecosystem, and purchases wafers.Intel’s work with Fortinet appears broader than that model. It combines:
- Fortinet’s specialized security-processor architecture.
- Intel’s semiconductor design capabilities.
- Back-end design assistance.
- Wafer manufacturing.
- Potential advanced packaging integration.
- Supply-chain diversification benefits.
Many companies need specialized chips for networking, cybersecurity, industrial systems, automotive platforms, cloud infrastructure, and edge AI appliances. Yet not all of them have the internal semiconductor engineering resources of Apple, Nvidia, AMD, Qualcomm, or large hyperscale cloud providers.
Intel can potentially serve these customers with a more complete package: processor IP where appropriate, engineering support, manufacturing, packaging, validation expertise, and supply-chain capability. In this model, Intel competes not only with Taiwan Semiconductor Manufacturing Co. but also with custom-silicon specialists such as Broadcom and Marvell.
That could create a valuable niche. Intel does not need to win every high-profile smartphone processor or AI GPU design to build a viable external business. It can win purpose-built chips where customers prioritize performance, security, integration, domestic or geographically diversified manufacturing, and engineering support.
Why Fortinet Matters to Enterprise IT
Fortinet’s products sit at the center of enterprise network security, firewall deployment, secure networking, and threat-protection infrastructure. A new security processor can influence the performance and capabilities of appliances that inspect traffic, enforce policies, handle encrypted connections, and support increasingly demanding workloads.As businesses deploy more encrypted traffic, cloud connectivity, AI-driven threat detection, and distributed branch environments, security hardware needs to process more data without becoming a bottleneck. Custom silicon is one way to improve efficiency compared with relying entirely on general-purpose CPUs.
For enterprise Windows environments, that can have indirect but important consequences. The security infrastructure protecting Windows endpoints, Active Directory environments, hybrid cloud resources, and remote-access deployments increasingly depends on high-throughput, specialized networking hardware.
A stronger Fortinet ASIC roadmap could improve appliance performance and efficiency. Intel’s participation gives the project a potential supply-chain angle as well, which may appeal to organizations that want more manufacturing resilience in critical infrastructure.
Still, the Fortinet win comes with limits. Intel has not disclosed the SP6 production schedule, expected volumes, packaging technology, fab location, financial terms, or date of volume manufacturing. Until SP6 reaches meaningful production, it cannot be treated as a major contributor to Intel Foundry revenue.
It is a promising proof point, not yet a revenue-scale breakthrough.
Intel 4 Is Not the Ultimate Foundry Test
Fortinet’s use of the Intel 4 process is strategically valuable, but it does not fully validate Intel’s most important external-foundry roadmap.Intel 4 is a proven production process with a track record inside Intel’s own product portfolio. That reduces technical risk for a customer such as Fortinet and gives Intel an opportunity to demonstrate commercial execution with an external partner.
However, the future of Intel Foundry depends more heavily on whether customers choose its newer nodes, particularly Intel 18A-P and Intel 14A.
The Significance of Intel 18A-P
Intel 18A-P is a performance-enhanced extension of Intel 18A. It is intended to offer additional performance and power benefits while building on the underlying 18A platform.The distinction matters because foundry customers do not choose a process node based only on a transistor roadmap slide. They evaluate a broad set of criteria:
- Performance per watt.
- Transistor density.
- Yield maturity.
- Design-rule stability.
- Process-design kit quality.
- Third-party IP availability.
- Memory compiler support.
- Timing libraries.
- Packaging compatibility.
- Cost per wafer.
- Geographic manufacturing options.
- Confidence in delivery schedules.
Intel has said that 18A-P has entered risk production. That is an important technical milestone, but risk production is not the same as broad commercial volume. It is a stage in which process technology is tested, refined, qualified, and prepared for customer designs.
Intel 14A Is the Bigger Strategic Bet
The next major test is Intel 14A, which Intel now plans to bring into risk production for internal products in the second half of 2027 before a high-volume ramp in 2028.Intel’s commitment to 14A high-volume manufacturing is a notable signal. Advanced-node development is extremely expensive, and companies do not commit to that scale of investment lightly. Intel’s management has pointed to improving defect density, transistor performance, process-design kits, and intellectual-property progress as reasons for confidence.
Yet the wording around 14A remains revealing. Intel is committing to ramp the node for its own products, while a major named external production customer has not been announced.
That is the gap Intel must close.
A major outside customer choosing Intel 14A for a high-volume product would carry far more weight than a vague statement of customer engagement. It would demonstrate that a company with other manufacturing choices believes Intel can meet strict requirements for confidentiality, process maturity, design enablement, capacity, packaging, and long-term support.
Until then, Intel 14A remains a technically ambitious roadmap with commercial potential rather than a proven external-foundry platform.
The Central Challenge: Trust When Capacity Gets Tight
Intel’s biggest foundry obstacle may not be transistor performance. It may be organizational trust.A pure-play foundry, by design, does not compete with its customers in major end markets. Its role is to manufacture chips for others. Intel is different. It owns major processor product lines in client computing, servers, graphics, networking, and AI infrastructure.
That creates a natural concern for outside customers: what happens when Intel’s own products need the same scarce factory capacity, engineering resources, packaging lines, substrates, and advanced manufacturing equipment?
Why External Customers Need Clear Assurances
A chip company committing a flagship design to a foundry is making a long-term bet. The project can take years from architecture planning through tape-out, qualification, production, and product launch.During that period, a customer needs assurance that:
- Its intellectual property will remain protected.
- Its product roadmap will not be exposed to Intel’s competing business units.
- Its wafers will receive dependable allocation during shortages.
- Its engineering issues will receive prompt attention.
- Its packaging and test needs will be met on schedule.
- Its commercial relationship will remain stable through market cycles.
External customers need repeated evidence. They need to see Intel honor commitments through difficult periods, not only when factory capacity is plentiful.
The company’s current internal supply constraints make that test more immediate. Intel is expanding Intel 3 capacity for Granite Rapids server processors while increasing 18A output for new client and server products. Those internal ramps help foundry utilization, but they also highlight exactly why external customers may want contractual and operational protections.
A foundry customer does not want to be treated as overflow demand. It wants confidence that it is a strategic partner.
Capital Spending Signals Confidence, but Also Raises Stakes
Intel expects capital spending to exceed $20 billion, with investment set to rise further as it expands manufacturing tools, front-end capacity, advanced packaging, substrates, and memory-related supply.That spending is immense, but it reflects the economic reality of the semiconductor industry. Leading-edge manufacturing requires multibillion-dollar commitments years before a product reaches store shelves or data centers.
Intel has emphasized that it is matching investment to customer demand rather than building large amounts of speculative capacity. That is financially prudent. It also suggests management believes it has enough confidence in future demand to place major equipment orders.
Still, there is a critical question: how much of that demand comes from external foundry customers, and how much comes from Intel Products?
The company has not provided a detailed breakdown. That limits outside visibility into whether foundry investment is being driven primarily by a growing merchant business or by the need to supply Intel’s own processors.
The Positive Interpretation
The optimistic case is compelling:- Intel’s fabs are becoming more productive.
- Intel’s internal CPU demand is improving utilization.
- Intel 18A is scaling faster than expected.
- Advanced packaging demand is increasing.
- External customer engagement is progressing.
- Fortinet demonstrates that the company can close commercial partnerships.
- Larger capacity investments could prepare Intel for substantial future foundry demand.
The Risk Case
The skeptical interpretation is equally important:- Intel may still be relying heavily on internal transfers rather than merchant-foundry revenue.
- A large capital program can strain cash flow if external customer ramps arrive slowly.
- Intel 14A remains expensive and commercially unproven with outside customers.
- Advanced packaging capacity could become another bottleneck.
- External customers may hesitate until Intel demonstrates stronger separation from its product businesses.
- Existing leading-edge foundry competitors have deeply established design ecosystems and customer relationships.
Intel Does Not Need to “Beat” TSMC Immediately
It is tempting to frame Intel Foundry as an immediate challenge to TSMC. That framing is too simplistic.TSMC has an enormous lead in merchant-foundry scale, customer trust, ecosystem depth, and advanced-node volume. It serves a broad collection of the world’s most demanding semiconductor companies and has built decades of credibility as a manufacturing partner.
Intel is not likely to displace that position quickly. Nor does it need to do so for Intel Foundry to succeed.
Supply-Chain Optionality Is a Real Market Opportunity
The more realistic near-term opportunity is optionality.A customer may not move an entire flagship design away from its incumbent foundry. But it may want:
- A second manufacturing source.
- A geographically diversified production option.
- A dedicated source for selected product lines.
- An advanced packaging partner.
- A custom-silicon partner for a specialized ASIC.
- A way to reduce dependency on a single supply chain.
- A potential alternative during capacity shortages.
Intel’s U.S. and European manufacturing footprint may become a commercial advantage where supply-chain resilience and regional manufacturing are strategic priorities.
But optionality is an intermediate stage, not the final goal. Intel’s defining external-foundry milestone will arrive when a major company chooses Intel as the primary production source for a high-value chip rather than simply as a hedge against capacity risk elsewhere.
What This Means for Windows PCs and the Broader Hardware Market
Intel Foundry may sound like a business-to-business story, but its impact reaches the Windows ecosystem.A healthier Intel manufacturing operation could improve the company’s ability to deliver competitive processors for Windows laptops, desktops, workstations, and enterprise systems. Better yields and improved wafer costs can support higher-volume availability and potentially give Intel more flexibility in product segmentation and pricing.
The effects may extend across several areas:
- Windows laptops: More efficient process nodes can help Intel build chips with stronger battery life, integrated graphics, and AI acceleration.
- Desktop PCs: Improved manufacturing economics may support better product availability across mainstream and enthusiast platforms.
- Enterprise PCs: Stable CPU supply matters for large fleet refreshes, especially as organizations modernize hardware for AI features and operating-system transitions.
- Servers and cloud systems: Intel’s manufacturing capacity supports Xeon platforms used in enterprise infrastructure and Windows Server deployments.
- Security appliances: Fortinet’s SP6 effort could contribute to more capable hardware used to protect corporate networks and remote-access environments.
- Supply resilience: More advanced chip manufacturing outside a single geographic concentration could reduce risk across the PC and enterprise hardware supply chain.
But Intel Foundry becoming more effective gives the broader Windows hardware ecosystem another potential source of advanced manufacturing capacity. That matters in an era when chip availability can shape everything from laptop launch timing to data-center expansion plans.
The Next Proof Points for Intel Foundry
Intel’s second-quarter results establish a stronger baseline. The company has shown it can reduce foundry losses, improve yields, increase output, and attract at least one meaningful commercial customer.The next phase requires more concrete evidence.
Key milestones to watch
- Fortinet SP6 tape-out and production timing.
- Whether Intel discloses meaningful production volumes for Fortinet.
- Additional named external customers for Intel 4, Intel 3, or Intel 18A-P.
- A major external commitment for Intel 14A.
- The maturity of Intel’s process-design kits and third-party IP ecosystem.
- Advanced packaging customer ramps, including EMIB-T deployments.
- Evidence that Intel can prioritize external customers through periods of tight capacity.
- Further reduction in Intel Foundry operating losses as external revenue grows.
The company needs a portfolio of customers across different industries, process nodes, product categories, and production volumes. It also needs those customers to return for future generations of silicon. Repeat business is one of the clearest indicators that a foundry relationship has moved from experimentation to trust.
A Better Quarter, but the Hardest Work Is Still Ahead
Intel Foundry’s second-quarter results deserve more credit than a narrow focus on its remaining losses would suggest. Revenue is growing, operating losses are shrinking, manufacturing efficiency is improving, and Intel 18A production is scaling.The Fortinet agreement also represents a legitimate external customer win. It shows that Intel can combine its design, manufacturing, and packaging capabilities into a commercial offering for a company building specialized silicon.
Yet Intel’s core challenge remains unresolved. The company is still primarily a manufacturer for itself, and its external revenue remains modest relative to the scale required to support a global merchant-foundry ambition.
The real measure of success will not be whether Intel can improve utilization in fabs filled by its own processors. It will be whether outside chip companies choose Intel for major products, remain confident that their intellectual property and production priorities are protected, and commit enough volume to turn Intel Foundry into a durable independent business.
Intel has demonstrated that its manufacturing recovery is gaining momentum. The next breakthrough must be commercial: not simply proving that Intel can make better chips for Intel, but proving that other companies want Intel to make their most important chips too.
References
- Primary source: EE Times
Published: 2026-07-24T22:00:00+00:00
- Related coverage: nextwavesinsight.com
Intel Foundry and 18A: Production Milestone, No Major Customers
intel foundry entered high-volume 18A production in January 2026. The yield data is improving. The external customer pipeline is not.nextwavesinsight.com - Related coverage: intel.com
- Related coverage: tomshardware.com
Intel will cancel 14A and following nodes if it can't win a major external customer — move would cede leading-edge nodes to TSMC and Samsung | Tom's Hardware
Then exit the leading-edge process technologies completely.www.tomshardware.com
