Business & Industry

Nvidia Is Putting $3 Billion Into Meta to Make Muse Cheaper to Run

Nvidia will invest $3 billion in Meta's equity offering and expand a partnership around Muse, Meta's AI assistant, with Vera Rubin systems, Vera CPUs, Spectrum-X networking and BlueField-4 processors on the hardware list.

Daniel HarrisDaniel Harris
Heat: 1,660
Nvidia Is Putting $3 Billion Into Meta to Make Muse Cheaper to Run

Nvidia said on October 27 that it will put $3 billion into Meta's equity offering and widen a partnership built around Muse, Meta's AI assistant. The stated goal is cheaper training and cheaper inference on hardware Meta has already committed to buy.

Nvidia Is Putting $3 Billion Into Meta to Speed Up Muse

Nvidia will invest $3 billion in Meta's equity offering, the chipmaker said on October 27. At the same time, the two companies are expanding work around Muse, the assistant Meta has been building for consumers and the model family behind it.

Money and hardware tend to travel together in deals like this. Nvidia's stated goals are to improve how Muse trains, to speed up how it answers, and to bring down the cost of running those workloads.

Neither company published a schedule for the investment. Neither put a number on the savings it expects. Meta said its Muse models show solid benchmark results and rising consumer use, without sharing new revenue figures or a measured cost reduction.

What Meta's Muse Assistant Actually Is

Muse is the personal assistant Mark Zuckerberg has been promoting as Meta's answer to ChatGPT. It was developed internally under the codename Hatch. The pitch is an agent that finishes tasks across everyday apps instead of waiting to be asked one question at a time.

That means links into email, calendar, shopping and payments, with the user able to pull access back whenever they want. Whether it behaves that way in daily use is another question, and one a press release can't answer.

For Nvidia, Muse matters as a workload. A consumer assistant that millions of people open every morning generates a huge amount of inference, and inference is where the bills pile up once a model is trained. Training is a capital project with a finish line. Serving users never stops.

Vera Rubin, Vera CPU and the Rest of the Hardware List

Meta plans to deploy more of Nvidia's next-generation Vera Rubin systems, along with Vera CPUs. The networking side adds Spectrum-X Ethernet and BlueField-4 processors.

The names matter less than the category. Vera Rubin is the rack-scale platform Nvidia is positioning for the next round of training and serving. Spectrum-X is how the company wants traffic to move between GPUs inside a data center. BlueField-4 takes networking and storage work off the main processor so it can spend its time on math.

The two companies are also discussing Nvidia's OpenShell architecture and Sentry monitoring technology, which are meant to put boundaries around autonomous AI systems. That's a governance question wearing a hardware jacket, and it's the part of the announcement with the least detail attached.

Why Nvidia Investing in a Customer Draws Fire

Nvidia isn't only selling chips to Meta now. It's also a shareholder, and that's where the criticism starts.

The pattern has a name in the current debate: circular financing. A supplier puts money into a customer, the customer spends it on the supplier's products, and the supplier books revenue. Each step can be described as ordinary business. The sum of the steps is what makes people uneasy.

Supporters say this is how large infrastructure buildouts have always been funded. Companies with capital help the companies that need it, and the market prices the risk. If Meta's AI business grows into the capacity it's buying, $3 billion will look cheap in hindsight.

Critics say the loop blurs how much independent demand exists. If part of the money that pays for chips comes from the company selling the chips, the revenue line is less clean than it looks.

There's no way to settle that argument today. What's checkable is disclosure. Nvidia named the amount in public, and Meta has to report the equity sale to regulators.

The $97.5 Billion Offering Behind the Deal

Reuters reported that Meta has expanded its planned equity financing to roughly $97.5 billion, with Nvidia's $3 billion as one slice. That figure comes from reporting rather than the joint statement.

Put the size of the raise next to the size of the check. Three billion is a small share of a raise that large, which suggests Nvidia is buying a seat at the table more than it's funding the buildout.

Meta's spending plans explain why the raise exists. Data centers, power contracts and chips are front-loaded costs, while revenue from AI products arrives later. Equity is one way to carry that gap without leaning entirely on debt.

The deal also fits a pattern across the industry this year. Chip suppliers have been taking stakes in the companies that buy from them, and cloud providers have been signing long contracts with the model labs that run on their servers. Balance sheets are getting tangled on purpose, and the accounting treatment matters as much as the strategy.

What to Watch in the Nvidia Meta Muse Deal

Two numbers will tell you whether this partnership changes anything. Watch for a cost-per-query figure from Meta, since cheaper to run is the whole promise. And watch whether Muse reaches general availability, because inference savings only show up at scale.

For anyone using Meta's apps, the effect is indirect. A cheaper assistant is one that can be given away to more people, and that's the version of this deal worth caring about. It depends on results nobody has published yet.

The hardware list gives one clue about timing. Vera Rubin systems aren't shipping in volume yet, so the cost improvements tied to them land in future quarters, not this one. Until then, the partnership is a promise with a purchase order attached.

Share This Story

Sources

Related AI News

A wireframe globe inside a broken ring, with a drone silhouette casting a targeting cone onto it
Business & Industry

The UN Spent a Week on Killer Robots and Agreed to Keep Talking

Diplomats left Geneva with permission to keep talking, not a rulebook. The two-year mandate that came out of the week is a starting gun for proposals, not a deadline for a ban.

Heat: 1,180
Rising bar columns crossing a flat dashed target line, with small particles drifting upward
Business & Industry

Meta Is Rethinking Its 2030 Net-Zero Pledge Because of AI

Meta's own sustainability report says the 2030 date is under review. The numbers behind that sentence are less ambiguous than the sentence itself.

Heat: 1,280
A falling candlestick line crossing a chip wafer, with a broken floor line beneath it
Business & Industry

AI Doubts Just Cost the Nasdaq 10.4% in a Week

The index lost 10.4% in five sessions and sits about 20% below its October peak. This wasn't one company's bad Friday, and that's the part worth understanding.

Heat: 1,950
A falling candlestick chart with a magnifying glass over ledger lines below it
Business & Industry

Nvidia Lost $900 Billion in a Week and the SEC Is Now Reading the Books

The stock fell about 14% on Friday and has now dropped for seven straight weeks. The SEC examination reported the same day is about the books, not about the founder.

Heat: 1,950