Intel Commits to 14A High-Volume Production in 2028

Intel has officially committed to high-volume manufacturing of its most advanced chipmaking node, 14A, in 2028 — a decision that represents a dramatic reversal from just three months earlier, when CEO Lip-Bu Tan warned the node could be delayed or even abandoned without sufficient customer demand.

The commitment came during Intel's Q2 2026 earnings call on July 23, alongside a strong beat-and-raise quarter that provided the financial foundation for the accelerated timeline. Here's a comprehensive breakdown of the key details.

Timeline Shift: One Year Faster

Intel pulled its 14A schedule forward by roughly one year. Risk production for internal products is now planned for the second half of 2027, with high-volume manufacturing (HVM) to begin in 2028 .

That timeline puts Intel's 1.4nm-class node in line with TSMC's A14 process . Just three months earlier, in Intel's Q1 2026 earnings call, Tan had warned that 14A could be delayed or abandoned without sufficient customer demand — a stance that has now fully reversed .

"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 fully commit to high volume ramp in 2028," Tan said on the earnings call .

Technical Progress: Better Than 18A at the Same Stage

Intel provided several technical updates on the 14A node's development:

• Yield ramp: 14A is in an early yield phase with a reported defect density (D0) of approximately 0.5, according to Morgan Stanley research. Early-stage ramp characteristics are stronger than Intel's previous 18A node .
• High-NA EUV: Intel has confirmed that 14A is being designed to incorporate an industry-first use of high-NA EUV lithography in high-volume semiconductor production . This is a critical competitive advantage over TSMC, which is not expected to deploy high-NA EUV until later nodes.
• PDK release: Intel has issued a 0.5 process design kit (PDK) for test chips and is working toward a 0.9 PDK, which CEO Lip-Bu Tan has called the "Holy Grail" of the 14A process. The 0.9 PDK is expected to be available to external customers by October 2026 .
• Backside power: The node will include an enhanced version of Intel's backside power delivery architecture, building on the PowerVia technology first deployed in 18A .
• Performance confidence: Tan emphasized that 14A is competitive across five core dimensions: performance, power, density, cost, and delivery timeline .

In a notable contrast with its recent foundry history, Intel says 14A is developing better than 18A did at the same stage, particularly on defect density and transistor performance metrics .

Capital Expenditure: $20 Billion and Rising

Intel raised its 2026 capital expenditure forecast from $18 billion to $20 billion, driven by surging AI data center demand and the need to build out 14A capacity .

CFO Dave Zinsner had previously signaled restraint on 14A capacity spending until customers were secured, telling analysts, "Our restraint is with 14A because it is closely tied to foundry clients. It doesn't make sense to significantly expand capacity until we confirm we have customers who will meet that demand" .

However, during the Q2 call, Tan indicated that the company is now "fully committed" and will layer expenses as customer traction materializes .

The capex increase reflects Intel's strategy to spend more on manufacturing tools — particularly for high-NA EUV — while spending less on new cleanroom expansions in the near term .

External Customer Wins: Tesla Confirmed, Apple Unconfirmed

Tesla (confirmed)

On Tesla's April 22 earnings call, Elon Musk disclosed that Tesla intends to use Intel's 14A process for chip production at its Terafab AI chip facility in Austin, making Tesla the first named external customer for Intel's most advanced node .

The deal covers Tesla's custom AI6 silicon and marks Intel Foundry's biggest win to date . Musk described Intel as "a key partner in TeraFab's expansion," and the partnership also extends to SpaceX and xAI . Intel disclosed the Terafab alliance (including SpaceX and xAI) during its Q1 2026 earnings call .

The Tesla win is significant because, just weeks earlier, Intel had warned it might exit the foundry business entirely without an external 14A commitment .

Apple (unconfirmed / preliminary)

Multiple reports point to a potential Apple deal, but no firm confirmation exists from either company:

• The Wall Street Journal reported on May 8, 2026, that Apple and Intel reached a preliminary foundry agreement for Intel to manufacture Apple-designed chips .
• Separately, TrendForce reported that Apple is evaluating Intel's 18A-P process (not 14A) for future M-series chips .
• In April, President Donald Trump publicly stated that Apple had agreed to make processors with Intel, but neither Apple nor Intel has confirmed the deal .
• Current evidence suggests any Apple involvement likely targets 18A-P in the near term, not 14A. Apple is reportedly evaluating 18A-P for next-generation M-series chips .

The key distinction: if Apple does move to Intel, it will likely start with 18A-P (the performance-enhanced variant of the 1.8nm node), not the more advanced 14A node.

Q2 2026 Earnings Context: The Financial Foundation

Intel's strong Q2 2026 results provided the financial basis for the 14A commitment:

• Revenue rose 25% year-over-year, driven by AI data center demand for its server CPUs .
• Q3 guidance came in at $15.8–$16.8 billion, well above analyst expectations .
• Production on Intel 7, Intel 3, and 18A all exceeded internal targets; 18A-P entered risk production .
• The company signed three-to-five-year agreements for data center CPUs and XPUs .

Tan stated that "extraordinary demand" for Intel's own CPUs, combined with encouraging external customer progress, gave the company confidence to fully fund and commit to 14A HVM .

The improved financial position is critical because 14A requires massive capital investment. As Tan noted earlier in 2026, "The increase in capital cost at Intel 14A make it clear that we need both Intel Products and a meaningful external customer to drive acceptable returns on our deployed capital" .

What's at Stake

Intel's 14A commitment comes with significant pressure. The company now has roughly two years to prove customers will place enough orders to justify the investment .

The timeline also intersects with U.S. policy deadlines: the enhanced 35% advanced manufacturing investment credit signed into law applies only to fab construction that begins before December 31, 2026 — meaning projects that break ground in 2027 receive no tax credit .

With Tesla secured and Apple potentially evaluating 18A-P rather than 14A, Intel's foundry strategy still hinges on winning additional high-volume customers for its most advanced node. Tan has indicated that multiple external customers are lining up as the company finalizes its 0.9 PDK .