Intel Raises $20 Billion to Fuel Its AI Chip Manufacturing Push

Intel raised $20 billion in a single stock sale this week, upsized from an original $15 billion plan, priced at $95 a share, with the deal closing today, August 12, 2026 — money the company says will bankroll new fabs, advanced packaging, and the 14A chipmaking process it’s betting its entire foundry comeback on. I’ve watched Intel talk about “returning to leadership” for the better part of three years now, and this is the first time the number attached to that promise has actually been bigger than the skepticism around it.

The offering itself is straightforward on paper. Intel priced 210,526,315 shares of common stock at $95 apiece, a little over a 2.6% discount to where the stock had been trading, and expects to pocket roughly $19.7 billion after underwriting fees. J.P. Morgan, Goldman Sachs, Morgan Stanley, and Citigroup ran the book, with a long list of additional banks — Barclays, BofA, Deutsche Bank, Mizuho, Wells Fargo, and others — along for the ride. Underwriters also got a 30-day option to buy another 31.5 million shares if demand holds. According to Bloomberg’s reporting on the deal, that demand was real: the sale reportedly drew more than $100 billion in orders, five times what Intel was actually selling.

Why Intel needed the cash now

Intel didn’t raise $20 billion because it was short on runway. It raised it because the AI buildout has turned semiconductor manufacturing into one of the most capital-hungry businesses on the planet, and Intel is trying to do something almost no company has pulled off from behind: catch up to TSMC while building out a contract-manufacturing business at the same time.

The money is earmarked for new fabrication capacity, advanced packaging lines, and — the part that matters most — scaling up 14A, Intel’s next-generation process node, toward high-volume production by 2028. Intel is also pouring roughly €5 billion into expanding manufacturing capacity in Ireland, one of its long-standing production hubs outside the U.S. This isn’t a company hoarding cash for a rainy day; it’s a company trying to buy its way back into a race it fell behind in during the mid-2020s, when Apple, Nvidia, and Qualcomm all leaned on TSMC instead.

What’s changed is the customer list. Intel has reportedly signed Tesla as a 14A customer — a genuinely notable win, given Tesla’s own chip ambitions are tied up in its Texas “Terafab” plans with SpaceX (I covered that deal here, and it’s worth noting nobody has fully confirmed who operates that fab either — this industry runs on overlapping bets right now). There’s also persistent chatter that Apple could eventually route some processor manufacturing through Intel’s foundry, though that one is still speculation, not a signed contract, and I’m not going to write it as more than that.

The stock’s wild 2026

Context matters here. Intel’s stock has nearly tripled since the start of 2026 — a run that has badly outpaced AMD, Nvidia, and the Philadelphia Semiconductor Index, which itself is only up around 75% this year. That’s an enormous re-rating for a company that spent 2023 and 2024 being written off as a legacy chipmaker that missed the AI wave entirely. Shares actually dipped more than 4% the Monday before the offering was announced — investors bracing for dilution, as they usually do — before recovering once the pricing and demand numbers came in.

Russ Mould, an investment director quoted on the deal, put it about as plainly as it can be put: as “a capital-intensive business,” it makes “perfect sense for Intel to raise money” while investors are this willing to hand it over. Strike while the iron’s hot, in other words, before sentiment on Intel’s turnaround cools off again.

How this fits the bigger AI-hardware picture

Intel’s raise doesn’t happen in a vacuum. It’s the same week Nvidia lined up $500 billion in financing commitments from six of Wall Street’s biggest asset managers to fund AI infrastructure buildouts for its own customers, and it follows SK Hynix’s blockbuster Nasdaq debut back in July, which I wrote about at the time — another chipmaker tapping public markets specifically because AI demand has outrun what internal cash flow can fund. The pattern across the whole sector right now is the same: whoever can manufacture at the leading edge is raising money as fast as investors will give it to them, because falling behind on a process node in 2026 is a multi-year, multi-billion-dollar mistake to correct.

For Intel specifically, the stakes are almost existential. The company’s foundry business — the unit that manufactures chips for other companies, the way TSMC does — has bled money for years while Intel tried to prove it could hit the same transistor density and yields as its Taiwanese rival. 14A is the node where Intel has said, repeatedly, that it expects to reach parity or better. $20 billion buys a lot of runway toward that goal, but it doesn’t buy certainty. Process nodes slip. Yields disappoint. Intel has been burned by both before.

What I’m watching next

I’ll be watching three things over the next few quarters: whether 14A stays on track for 2028 high-volume production, whether any more marquee customers beyond Tesla actually sign on the dotted line, and whether Intel’s stock can hold this valuation once the “raised a lot of money” headline fades and the market goes back to asking “are the chips actually shipping.” A $20 billion raise is a vote of confidence from Wall Street. It is not, on its own, a finished comeback.

Quick reference: Intel’s $20B offering at a glance

Frequently Asked Questions

Why did Intel raise $20 billion instead of $15 billion?

Intel originally planned a $15 billion offering, but investor demand came in so strong — reportedly north of $100 billion in orders — that the company and its underwriters upsized the deal to $20 billion to capture more of that appetite while it was there.

What is Intel’s 14A process, and why does it matter?

14A is Intel’s next-generation chip manufacturing node, the successor to its current leading-edge processes. It’s central to Intel’s plan to compete directly with TSMC for contract manufacturing customers, and the company is targeting high-volume production on it by 2028.

Is Tesla actually manufacturing chips with Intel now?

Tesla has reportedly been secured as a customer for Intel’s 14A process, though full production details haven’t been laid out publicly. It’s a separate relationship from Tesla and SpaceX’s own planned “Terafab” chip plant in Texas.

Does this offering dilute existing Intel shareholders?

Yes — issuing roughly 210 million new shares increases Intel’s total share count, which is standard dilution for a stock offering. That’s also why the stock dipped in the days before the deal was priced, even though it had been on a strong run for most of 2026.