Intel vs TSMC stock: Musk redraws the Terafab bet

Elon Musk narrowed TSMC’s proposed role in Terafab on Wednesday, saying his companies intend to build and run the chip factory themselves, while TSMC might sublease part of it. His October 7 statement changes the question for investors comparing Intel and TSMC stock: who can turn involvement in the project into paid manufacturing work?

Intel already has a stated Terafab partnership. Its prepared earnings remarks in April described exploring improvements to semiconductor manufacturing with SpaceX, xAI and Tesla. But an intended operator, a technology partnership and a customer order are different things. Wednesday’s clarification does not supply a new order value or an earnings contribution for either Intel or TSMC.

In the U.S. session, TECHi’s quote feed recorded Intel at $113.85, up 1.20%, at 10:50 a.m. EDT, and TSMC’s New York-listed ADR at $475.58, down 1.39%, at 10:26 a.m. EDT. These October 7 observations may be delayed; their timestamps differ, so they are not a synchronized return comparison. The moves alone cannot establish how much traders attributed to Terafab.

What Musk’s Terafab clarification changes

“We will build and run the fab,” Musk wrote. He described a possible TSMC sublease, rather than announcing that TSMC would operate the whole facility. That is his stated intention; the post does not establish that TSMC has accepted the arrangement.

For shareholders, the missing commercial details matter more than the factory’s name. A sublease would govern the use of factory space; it would not by itself define chip orders or manufacturing revenue. Process technology, wafer manufacturing and advanced packaging could create different revenue streams. None should be entered into an earnings model simply because a company appears in a discussion about the same project.

The immediate comparison therefore favors evidence over presumed exclusivity. Intel’s April remarks confirm a partnership, but do not make Wednesday’s post a new Intel contract. TSMC’s possible presence does not establish an agreed order, either. The questions to ask are what work each company would perform, who funds the capacity and when a customer must pay.

Article Brief

What changes for Intel and TSMC investors

3 Points18s Read

• 01The news–Musk says his companies intend to run Terafab, with a possible TSMC sublease. This is not an announced TSMC acceptance or new order.
• 02The financial distinction–Only about 5.1% of Intel Foundry Q2 segment revenue came from external customers. TSMC already reports substantial dedicated-foundry earnings.
• 03The evidence to watch–Customer commitments, funding and manufacturing economics matter more than an assumed exclusive role.

Intel vs TSMC stock: compare the earnings base first

The latest completed second quarters show why the same Terafab headline can mean different things to these businesses. Intel’s July results cover the quarter ended June 27. TSMC’s results cover the quarter ended June 30.

Latest completed Q2: company-wide results

Reported metric	Intel Q2 2026	TSMC Q2 2026
Quarter ended	June 27, 2026	June 30, 2026
Company revenue, USD	$16.128 billion	$40.20 billion
Reported gross margin	40.4%	67.7%
Reported operating margin	11.1%	60.3%
Accounting basis	U.S. GAAP	TIFRS
Next Q3 results	October 29 after U.S. close	October 15

These are company-wide results, not an equal comparison between two stand-alone foundries. Intel combines chip products with manufacturing; its Foundry segment includes internal business. TSMC describes itself as a dedicated foundry. The accounts also use different standards: U.S. GAAP for Intel and TIFRS for TSMC. The table establishes today’s reported earnings base, rather than proving which stock offers the better return at its current price.

That distinction makes Intel more sensitive to evidence that its external manufacturing business can grow. TSMC already has substantial reported revenue and margins to defend. A proposed project could matter strategically to both without becoming equally important to their near-term earnings.

The 5.1% figure behind Intel’s foundry opportunity

Intel’s second-quarter Form 10-Q reports $5.765 billion of Foundry segment revenue, including $293 million from external customers. TECHi’s calculation is $293 million ÷ $5,765 million × 100 = 5.08%, or about 5.1%. The remaining segment revenue was internal Intel activity.

The filing also attributes the external revenue increase primarily to Altera’s transition to an external customer after its deconsolidation. Investors should not interpret that increase as equivalent evidence of a broad wave of newly won independent customers. Intel Foundry reported a $2.089 billion quarterly operating loss.

This is the clearest reason a commercially defined Terafab role could matter to Intel. External work could widen its customer base. Yet volume alone would not answer whether the work earns an acceptable return after development, equipment and manufacturing costs. Intel needs disclosed economics alongside technical progress.

For TSMC, accepting a limited role could still be rational if the commercial terms are attractive. Conversely, declining a role need not undermine its existing business. That judgment requires terms that Wednesday’s statement does not provide.

The next tests arrive before Terafab’s economics are known

TSMC’s official calendar schedules September sales for October 8 and third-quarter results for October 15. Its published Q3 outlook calls for $44.6 billion–$45.8 billion of revenue and a 65%–67% gross margin. Those disclosures can test demand and profitability in the business investors already own.

Intel confirmed on October 6 that it will report Q3 results on October 29 after the market closes. Its July outlook was $15.8 billion–$16.8 billion of revenue with a 41% GAAP gross margin. Investors can then check operating execution and whether the company gives more specific external-customer commitments.

TECHi’s earlier TSMC reporting explains the October sales and earnings sequence. Our Intel foundry analysis examines why manufacturing progress must translate into better economics. The new Terafab statement adds a question about roles; it does not replace either financial test.

What would change the comparison?

Intel’s stronger case would be a named commercial scope, committed customer volume, funding terms and a credible path to lower foundry losses. Investors should also examine capital requirements and any share issuance needed to finance expansion. A partnership can be valuable while still taking years to benefit each existing share.

TSMC’s stronger case would be continued execution against its revenue and margin outlook, with any Terafab involvement priced on terms that justify the resources required. Its existing earnings base offers clearer operating evidence, but it does not remove valuation, capacity-investment or currency risks.

The counterargument to preferring TSMC’s established earnings is that Intel could benefit disproportionately from a successful external foundry expansion. The counterargument to buying Intel on the headline is that strategic relevance can arrive long before profitable revenue. Neither argument establishes a short-term price target.

For this Intel-versus-TSMC decision, the next meaningful disclosure is one that turns a proposed role into a financial commitment. Until then, treat Terafab as an unresolved opportunity and use the October results to judge the operating businesses underneath it.