For the better part of a decade, one company has manufactured the world’s most advanced chips, and everyone else has waited in line. This week, that queue sprouted a new branch.

On Wednesday, Tesla CEO Elon Musk confirmed plans to use Intel’s next-generation 14A manufacturing process at Terafab, the massive AI chip complex he has proposed building in Austin. The announcement makes Tesla the first major external customer for Intel’s most advanced manufacturing technology — and gives Intel’s foundry ambitions their first credible foothold against TSMC.

Intel CEO Lip Bu Tan had publicly warned that without a major external customer for 14A, the company would exit the chip manufacturing business altogether. In an Intel News video on X earlier this year, Tan struck a very different tone: “We are going big time into 14A.”

Now we know why.

Breaking the Monopoly

TSMC fabricates virtually all of the world’s bleeding-edge logic chips. Apple, Nvidia, AMD, Qualcomm — all are TSMC clients. Intel Foundry, despite billions in investment and years of effort, has failed to land a single major external customer for its current leading-edge 18A node in meaningful volumes. The 18A process will be consumed primarily by Intel’s own products, supplemented by modest orders from Microsoft and the US Department of Defense.

The 14A node, expected to reach risk production around 2028 and volume manufacturing in 2029, represents a significant technical leap. It introduces second-generation RibbonFET gate-all-around transistors, a refined backside power delivery system called PowerDirect, and Turbo Cells that boost critical timing paths for speed. According to Tom’s Hardware, the node builds directly on lessons from 18A — the process currently producing Intel’s Panther Lake processors.

But the specs matter less than the signal. Ben Bajarin, head of Creative Strategies, told Reuters that Intel’s 14A technology “could turn out to be a bigger deal for Intel than folks thought,” calling Tesla “a great first non-Intel customer” that “will definitely have scale.”

Jay Goldberg, an analyst at Seaport Research Partners, was more direct: “Having a customer is more important than the timing.”

The Terafab Unknowns

Tesla’s commitment arrives wrapped in characteristic Musk uncertainty. The Terafab project — a joint venture between SpaceX and Tesla — envisions two advanced chip factories: one for vehicles and humanoid robots, another for space data centers. Musk has claimed the facility will eventually produce one terawatt of annual computing capacity, compared with roughly half a terawatt currently generated across the entire United States.

The full scope of that vision would cost between $5 trillion and $13 trillion in capital expenditure, according to Bernstein estimates. Operational details — who funds the chipmaking equipment, who runs the factories, when production begins — remain largely undisclosed. WCCFTech reports that Terafab is targeting a 2029 operational start with approximately 3,000 wafers per month in the pilot phase.

Goldberg noted that Musk’s loftiest volume projections may not materialize, but Tesla’s existing businesses alone would represent meaningful demand for Intel’s foundry. “It’s not equivalent to Apple or Nvidia” in chip volumes, Goldberg told Reuters. “But it’s a real customer. It can be real volumes.”

The Capacity Catch

The deal gives Intel something its foundry has never possessed: external validation at the leading edge. But it also surfaces an uncomfortable reality. Intel’s current capital expenditure plan does not include investments in 14A capacity for third-party clients, according to Tom’s Hardware. John Pitzer, Intel’s vice president of corporate planning, told an RBC Capital Markets conference in November that winning a 14A customer would require the company to “layer on expenses well ahead of getting revenue,” pushing out Intel Foundry’s breakeven timeline.

Most investors, Pitzer suggested, would accept that delay as proof that Intel can “actually stand up an external foundry.” The market’s initial reaction was cautiously positive: Intel shares rose 3.6 percent in extended trading following Musk’s announcement. Intel declined to comment on the remarks, according to Reuters.

Musk framed the partnership as a timing play. “Given that by the time Terafab scales up, 14A will be probably fairly mature or ready for prime time,” he said during Tesla’s Q1 2026 earnings call. “14A seems like the right move, and we have a great relationship with Intel.”

For Intel, the question is no longer whether anyone will buy what it builds at the leading edge. It’s whether the company can scale fast enough — and invest boldly enough — to make that purchase matter.

Sources