Nvidia Recruits Six Wall Street Firms to Raise $500 Billion for AI Buildout

Today’s Focus

Nvidia said it has brought together six of the largest names in finance to assemble a pool of more than $500 billion (£370 billion) aimed at building out artificial intelligence infrastructure, according to reporting from the BBC and The Guardian.

The chipmaker signed memorandums of understanding with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR. The arrangements set up financing platforms designed to let outside investors put money into AI “compute,” the industry shorthand for the hardware and processing capacity that runs AI systems.

Chief executive Jensen Huang framed the effort as a first-of-its-kind move to treat that computing capacity as an investable asset class. “In AI, compute is revenue,” Huang said, according to the BBC, adding that the company was assembling long-term capital providers to underwrite AI infrastructure independently.

Huang wrote on X that Nvidia holds the option to backstop as much as $125 billion, roughly 25% of the potential deals, The Guardian reported. The money is set to fund Nvidia’s own projects along with those of its partners, including new data centers and factories that manufacture the chips these systems depend on.

Nearly every major AI operation relies on Nvidia’s graphics processing units, among them Google, Amazon, Microsoft, Meta, OpenAI, Anthropic, Tesla and SpaceX. The Guardian put Nvidia’s market value at $5.3 trillion and noted combined AI spending by big technology firms is projected to top $730 billion this year.

KKR co-chief executives Joe Bae and Scott Nuttall said in a joint statement that computing power had turned into a core piece of infrastructure. “Delivery, not ambition, is the hard part,” they said, per the BBC.

The Debate

Supporters argue

Backers of the arrangement describe it as a way to match the enormous, long-lived cost of AI infrastructure with investors who think in decades rather than quarters.

KKR’s Bae and Nuttall argued that computing power now functions like roads, power grids and other critical infrastructure, and that the difficulty lies in actually delivering projects at scale. Their firm, they said, has learned that lesson from expanding into digital infrastructure.

Huang cast the platforms as a solution to scarcity, telling investors the structures would let customers reach limited computing capacity and construct the “AI factories” he says will run every industry and country. Pooling institutional money, supporters contend, spreads risk across major players rather than loading it onto any single tech balance sheet.

The involvement of Apollo, BlackRock, Blackstone, Brookfield and Goldman Sachs, they note, signals that sophisticated capital allocators see durable demand. With Nvidia offering to backstop up to $125 billion, supporters say the company is putting its own money behind the projects it is asking others to fund.

Critics argue

Skeptics question whether the deal reflects genuine demand or a financing structure built to keep a boom going.

The Guardian reported concern about the tie between the sky-high valuations of technology companies and the vast sums they must raise to sustain their ambitions. Critics point to that gap as a warning sign: if AI revenue fails to grow into the spending, the losses could ripple through pension funds and other institutions now buying in.

Some observers see Nvidia’s backstop offer as a red flag rather than reassurance. A chipmaker helping to underwrite purchases of its own chips, they argue, blurs the line between selling hardware and financing its customers, a pattern that has ended badly in past technology cycles.

Turning compute into an asset class, critics add, could concentrate financial exposure in a narrow set of firms and untested projects. When spending across the sector is on track to exceed $730 billion this year, they say, even confident investors should ask how much of that rests on debt.

What the experts say

Independent analysts have flagged the scale of AI capital spending and the uncertainty around returns.

The International Monetary Fund warned in its April 2025 Global Financial Stability Report that a sharp correction in AI-linked equities could tighten financial conditions, noting that a small number of technology firms account for an outsized share of major stock indexes. That concentration, the IMF said, raises the stakes if sentiment shifts.

Historical comparisons offer a mixed record. Economists at the Federal Reserve and elsewhere have drawn parallels to the late-1990s telecom buildout, when firms borrowed heavily to lay fiber-optic cable ahead of demand; much of that capacity sat unused for years, though it eventually powered the internet economy.

The MIT-affiliated research group behind the 2024 “GenAI Divide” report found that most enterprise AI pilots had not yet produced measurable returns, underscoring the gap between spending and realized profit. Whether the $730 billion in projected 2026 outlays pays off, researchers say, depends on adoption that has not yet been demonstrated at scale.

By the Numbers

$500 billion: the capital Nvidia says it aims to raise with its six Wall Street partners, per the BBC and The Guardian.

$125 billion: the amount, or about 25% of the deals, that Nvidia says it can backstop, according to Huang on X via The Guardian.

Six: the number of financial firms signing memorandums of understanding, Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR, per the BBC.

$5.3 trillion: Nvidia’s market value as reported by The Guardian.

$730 billion: projected combined AI spending by big technology companies this year, according to The Guardian.

April 2025: the IMF Global Financial Stability Report warned that concentrated AI-linked equities could amplify a market correction.