Intel confirms Nvidia’s $5B stake is done — 214.8 million shares at $23.28 — as the companies sell an x86-friendly path into Nvidia’s AI infrastructure
Samuel Boivin/NurPhoto via Getty Images
Intel $INTC just cashed Nvidia $NVDA’s $5 billion check. Intel says it completed the share sale to Nvidia on Dec. 26 after regulatory clearance earlier this month, turning a September announcement into settled cash and issued shares — and making Nvidia’s roughly 4% stake in Intel official.
Nvidia bought 214,776,632 shares at $23.28 apiece in a private placement, a round-number infusion that landed on Intel’s balance sheet like a vote of confidence and on its reputation like a defibrillator. For months, the deal lived in that corporate purgatory where everyone speaks in the future tense and the market politely pretends it can’t hear the lawyers. Now, the cash is in — and the shares exist.
The price has gotten more interesting with time, too. Nvidia’s $23.28-a-share buy-in is 36% below Intel’s prior Friday close — a reminder that when the market’s main character writes a check, it doesn’t overpay for the privilege.
Nvidia was off about 1.7% in early morning trading, and Intel was relatively flat (up just over 1%). That’s what happens when a headline becomes a filing. Intel got the money. Now: It has to make the partnership feel like hardware, not hope.
It’s a private placement, not a public-market nibble, and Intel’s earlier filing emphasized Nvidia wasn’t getting special governance or information rights beyond what shareholders already get. The deal was conditioned on the regulatory process, and the FTC’s early-termination notice for the transaction is dated Dec. 18 — early termination of the Hart-Scott-Rodino waiting period, the legal box-check that lets deals stop being “pending” and start being “done.”
“AI is driving a reinvention of every layer of the computing stack,” Huang said back in September at a press conference to discuss the news, which he called a “historic partnership” and a “great business opportunity” for both companies. Huang said the three architecture teams — working across the CPU architecture, as well as product lines for servers and PCs — have been developing this “fairly extensive” architecture for over a year. He added, “Today, we’re taking the next great step.”
You could feel how unthinkable this pairing once was by the way the Street talked about it. “This is truly like the Yankees and the Red Sox coming together to end their rivalry — the companies did not like each other whatsoever,” David Wagner, head of equity and a portfolio manager at Aptus Capital Advisors, wrote at the time. “[Using Intel’s fabs to make chips is] a massive step in the right direction for U.S. chip designers and breathes new life into a poorly run company for decades.”
This partnership didn’t happen in a vacuum. Earlier this fall, the federal government took the rare step of engineering a roughly 10% equity stake in Intel, converting lectures about industrial policy into a live experiment with an actual balance sheet. The politics were loud, the stakes louder, and the message unmistakable: In President Donald Trump’s America’s AI playbook — a CEO-style, results-first posture toward national champions — Intel is the corporate vessel for “made here” ambition. But Intel didn’t just need capital. It also needed a partner.
The Intel-Nvidia partnership pitch stays the same, and the omissions stay just as loud. Intel and Nvidia have said they’ll co-develop multiple generations of products for data centers and PCs, including Nvidia-custom x86 CPUs meant to slot into Nvidia’s AI infrastructure and PC system-on-chips that combine Intel CPU cores with Nvidia RTX GPU chiplets, stitched together with NVLink. Missing, deliberately, is any promise that Nvidia’s crown-jewel GPUs will move off TSMC $TSM or that Intel will suddenly become the default factory for Nvidia’s most important silicon. Neither company has said whether Nvidia would use Intel’s contract manufacturing for its chips.
The deal reads like influence and distribution, not a manufacturing surrender. A big check, not a merger.
For Intel, that’s still a win. It gets cash plus the kind of association that turns a long-suffering turnaround narrative into something procurement teams can cite without blushing. For Nvidia, this is a smoother path into the x86 universe without asking enterprises to torch their software estates — and a convenient way to look less like a solitary superpower while regulators and politicians keep peering at the AI stack like it’s critical infrastructure.
Nvidia’s check has cleared — now, Intel’s story has to.
Join 500,000+ readers who start their day with Quartz.
By subscribing, you agree to our Terms of Service and Privacy Policy.
