A Turnaround Taking Shape
Intel's foundry turnaround strategy is showing concrete signs of progress in 2026, as the company secures major customer interest and delivers on its most ambitious process node roadmap in decades. Under CEO Lip-Bu Tan, who took the helm in March 2025, Intel has separated its foundry business as an independent unit and is executing what many analysts describe as the most ambitious transformation in semiconductor history.
The stakes are extraordinary. Intel reported foundry losses exceeding $7 billion in 2023 and 2024, with cumulative losses from the foundry venture now reaching tens of billions. The company's Q1 2026 earnings showed progress: revenue of $13.6 billion beat consensus of $12.36 billion, and adjusted EPS of $0.29 far exceeded expectations of just $0.01. For Q2 2026, Intel guided revenue of $13.8 to $14.8 billion, ahead of analyst estimates.
The 14A Node: A Make-or-Break Milestone
Intel's 14A process — its 1.4-nanometer-class node — is emerging as the centerpiece of the foundry turnaround. The release of the 14A Process Design Kit (PDK) version 1.0 has been identified by UBS as a key catalyst for attracting external customers. The node will be Intel's first to deploy High-NA EUV lithography from ASML, alongside second-generation RibbonFET gate-all-around (GAA) transistors and PowerDirect, an upgraded version of the backside power delivery technology introduced with 18A.
CEO Tan declared in a recent briefing: "We are going big time into 14A. Stay tuned, we are going to see a lot of great momentum on the 14A in terms of yields and IP portfolio to serve the customer well." This represents a sharp reversal from July 2025, when Tan suggested Intel might pause or discontinue 14A development if it failed to secure external customers.
The equipment order data confirms Intel's commitment. TrendForce reported in April 2026 that Intel Foundry has ramped chipmaking equipment orders by more than 50% year-over-year since the start of 2026. Taiwanese equipment suppliers including E&R Engineering and KINIK Company are among the beneficiaries.
Major Customer Interest
Perhaps the most encouraging sign for Intel's foundry ambitions is the caliber of customers evaluating its technology. According to UBS and multiple industry sources, Apple, AMD, Nvidia, Google, and Broadcom are all in discussions with Intel Foundry for potential manufacturing deals.
Apple is rumored to be considering manufacturing some of its M-Series "Apple Silicon" laptop processors using the 18A-P node, potentially as early as 2027. Google may leverage Intel's EMIB and Foveros 3D advanced packaging technology for some of its TPU designs. Nvidia has explored Intel's 18A and 14A nodes as potential secondary sources for future GPU generations, a move that would diversify its manufacturing base beyond TSMC.
These discussions represent a significant shift in the foundry landscape. For years, leading-edge chip designers had only one viable option for advanced nodes: TSMC. Intel's emergence as a credible second source gives customers leverage in pricing negotiations and supply security. The proximity of Intel's US-based fabs is also a selling point for customers concerned about the geographic concentration of advanced manufacturing in Taiwan.
The US Government Stake
In an unusual development, the US government has taken a direct ownership stake in Intel, purchasing $8.9 billion in common stock at $20.47 per share, giving the government a 9.9% stake. Additionally, through the CHIPS Act, Intel has secured $8.5 billion in grants and $11 billion in loans — a total of $19.5 billion in government support that materially de-risks Intel's capital burden.
The government investment reflects the strategic importance of Intel's foundry ambitions for US national security. A domestically based advanced foundry is a cornerstone of Washington's semiconductor policy, reducing dependence on Asian manufacturing for critical AI and defense chips.
The Gaudi AI Accelerator
While the foundry business receives most of the attention, Intel's own AI accelerator product — the Gaudi series — is also evolving. Gaudi 3 has found a niche in cost-sensitive inference deployments, competing with Nvidia's lower-end accelerators and AMD's Instinct lineup. Intel has positioned Gaudi as a cost-effective alternative for enterprises that do not need the extreme performance of Nvidia's flagship products.
Gaudi 4, manufactured on Intel's own 18A process, is expected to bring competitive performance to the mid-range AI accelerator market. The tight integration between Intel's foundry technology and its own chip designs allows the company to optimize its products for its manufacturing processes, potentially offering cost or performance advantages over fabless competitors.
The Path to Profitability
Intel targets foundry profitability by fiscal year 2028, with free cash flow positivity expected in fiscal 2027. This timeline requires sustained customer wins and high utilization of the company's expensive fab capacity. Analysts estimate Intel needs to secure over $10 billion in external foundry revenue to reach break-even.
The 18A node's success in customer qualification is the near-term catalyst. If Apple, Nvidia, or other major customers commit to volume production on Intel's nodes, the foundry business could reach a tipping point where capacity utilization and learning curve improvements drive rapid margin expansion.
Challenges and Risks
Despite the encouraging progress, Intel faces substantial challenges. Process node execution must be flawless — any delay or yield issue would damage credibility with customers who have alternatives. The foundry business model is fundamentally different from Intel's traditional integrated device manufacturer (IDM) approach, requiring a customer service orientation and IP ecosystem that the company is still building.
Intel's balance sheet remains under pressure, with free cash flow deeply negative at approximately negative $5.5 billion due to $20+ billion in annual fab capital expenditure. The company cannot afford another misstep like the 7nm delay that plagued its previous leadership.
The Verdict
Intel's foundry turnaround remains a high-risk, high-reward bet. The company has the technology roadmap — with 18A and 14A offering credible alternatives to TSMC's N2 and A16 — and the government backing to weather years of investment. What it needs now is execution and customer commitment. If Intel can convert the current customer interest into binding contracts and volume production, the foundry business could justify the billions invested. If not, it will go down as one of the largest capital misallocations in semiconductor history.
For the industry, a successful Intel foundry would be transformative: a third major player in advanced manufacturing would increase supply chain resilience, improve pricing competition, and accelerate the pace of process node advancement. The next twelve months will be decisive.