Elon Musk has confirmed that Taiwan Semiconductor Manufacturing could join the Terafab joint venture as a secondary manufacturing partner, directly threatening the anchor-customer thesis behind Intel's 14A process node. Musk confirmed a report by Tim Culpan that TSMC could supply Terafab with additional manufacturing expertise and capacity, one day after Intel had officially joined the end-to-end chip production project in Texas.

The news lands at a delicate moment for Intel shareholders. The stock has climbed 163% since the start of April and now trades at 63 times forward earnings estimates — a valuation that leans heavily on the foundry momentum built over the past six months, much of it tied to Terafab interest.

What does TSMC's possible role change?

When Intel appeared to be Terafab's exclusive partner, the deal read as a massive vote of confidence in its next-generation manufacturing process. It would give Intel an anchor customer, let it scale production faster, and serve more external customers from a confirmed demand base. Musk's discussions with TSMC cast doubt on whether Intel can meet Terafab's needs on its own.

TSMC has clear commercial reasons to engage. Any agreement will likely include financial backing for new facilities in Texas, where the world's leading foundry is reportedly considering multiple chipmaking plants, each carrying a price tag of at least $20 billion — on top of its existing $265 billion U.S. investment plan. Terafab would also hand TSMC a major customer to keep capacity highly utilized.

The strategic payoff may matter more than the revenue. If TSMC wins the node for Terafab's production, Intel's foundry business faces a huge setback just as it works to establish credibility against its chief rival.

Why does Musk need two foundries?

Musk's ambitions require far more compute than either TSMC or Intel can currently supply. SpaceX plans to launch orbital data centers within a couple of years. Tesla aims to scale humanoid robot Optimus production to 1 million units per year within five years, and expects robotaxi expansion to lift auto output as well.

Terafab exists to vertically integrate chip production for those two companies — not to compete with TSMC, Intel, or other foundries for third-party business. That distinction matters: partners have little to lose by licensing technology and lending expertise to the venture.

How strong is Intel's foundry business today?

Intel's foundry business generated $5.8 billion last quarter, up 31% year over year — but $5.5 billion of that came from Intel's own chipmaking operations. Rising CPU demand tied to AI has made Intel a bigger customer of its own fabs. To sustain its manufacturing ambitions, it still needs outside customers.

The 14A pipeline shows real movement:

• Intel has pulled forward its 14A production timeline by about one year.
• It will deliver process design kit (PDK) 0.9 to potential customers this month, enabling prototype designs.
• Risk production starts next year, with volume production in 2028 — nearly matching TSMC's schedule for its comparable A14 node.

Does 14A still lead on performance?

Both companies' estimates previously suggested Intel would take the lead in performance and power efficiency at the next node. Intel Foundry CTO and operations officer Naga Chandrasekaran recently tempered that expectation, saying 14A will deliver performance within 5% of TSMC's A14 node — a comment that may have prompted Terafab to approach TSMC in the first place.

A split outcome remains plausible. Intel could stay Terafab's leading manufacturing partner on 14A while TSMC provides ancillary services and expertise — notably advanced packaging, where TSMC commands roughly 95% of the market.

For now, investors are repricing a stock whose rally assumed Intel had Terafab largely to itself. Whether Intel keeps the lead role, shares it, or loses the node to TSMC will shape the competitive dynamics of the next-generation foundry race through 2028.