Nvidia agrees framework with six Wall Street firms to raise $500 billion for AI
The memorandums are not final agreements, and Huang said Nvidia may cover up to a quarter of any single deal.

Nvidia shares have climbed back toward record territory this month, lifting the chipmaker's market value above $5.3 trillion and keeping it the world's most valuable company. The stock closed at $219.74 on Aug. 18, up about 18% for the year, as three forces converged on it: a $500 billion financing framework agreed with Wall Street, still-rising capital spending by the largest cloud companies, and a fresh wave of processors moving into production ahead of the company's next earnings report.
This is really the first time that technology chips have become an investable asset class.
Huang made the remark on Aug. 10, when Nvidia said it had signed memorandums of understanding with six large asset managers to mobilize more than $500 billion in outside capital for artificial intelligence infrastructure. Nvidia said the partnerships remain subject to execution of final agreements. The company wants to treat its processors much like commercial real estate, toll roads or other long-lived assets, the sort of holdings that lenders will readily underwrite.
A $500 billion bet on borrowed chips
The six firms are Apollo Global Management, BlackRock, Blackstone, Brookfield Asset Management, Goldman Sachs and KKR. Executives from all seven companies, including Nvidia, appeared together in a live joint interview with CNBC's Becky Quick, a rare arrangement that underscored how much each stands to gain from financing the buildout of AI data centers.
The structure lets Nvidia's customers borrow against graphics processors and the data centers that house them rather than paying the full cost upfront. By pulling in institutional credit, insurance money and private capital, the arrangement widens the pool of buyers who can afford Nvidia systems. Huang told CNBC that Nvidia may provide residual value support on up to a quarter of any single opportunity, assessed project by project.
Huang has described AI compute as revenue-generating assets that are productive, long-lived and transferable between customers, the traits he argues make the hardware suitable as collateral. If a processor can be borrowed against and moved from one operator to another, it begins to behave like the infrastructure banks have financed for decades.
The initial market response was counterintuitive. Nvidia shares fell as much as 3.2% during the session and closed down 2.9% on Aug. 10 at $217.55, erasing about $155 billion in value. Traders focused on the risk of circular financing, the worry that a chipmaker helping to fund its own customers' purchases could be masking softer underlying demand.
Hyperscalers keep raising their spending
Concerns about demand run against what the biggest cloud operators are budgeting. Guidance from Microsoft, Amazon, Alphabet and Meta adds up to roughly $720 billion to $745 billion of capital spending in 2026, about 77% more than the previous year, on figures compiled by Euronews. Much of that money flows toward the servers and networking gear Nvidia sells.
Expectations for 2027 have climbed faster still. Vivek Arya, Bank of America's semiconductor analyst, put hyperscaler capital spending above $860 billion in 2026 and on a path toward roughly $1.2 trillion in 2027. The revisions help explain why chip investors shrugged off the summer's wobble.
Goldman Sachs frames the trend over a longer horizon, estimating that spending on AI computing will grow from about $494 billion in 2026 to $1.13 trillion by 2031.
Nvidia's own order book reflects that appetite. The company reported a backlog exceeding $500 billion for its Blackwell and Rubin platforms, a separate figure from the financing framework. Its most recent results, for the fiscal first quarter ended April 26 and reported on May 20, showed record revenue of $81.6 billion and record data-center revenue of $75.2 billion.
The next generation of chips
The second pillar is product. Nvidia said on May 31 that its Vera Rubin platform had ramped into full production, with shipments due in the fall, positioning it for so-called agentic AI, where a single user request can trigger many internal model calls, retrieval steps and checks. Rubin follows Blackwell, which moved the company into rack-scale systems.
Huang delivered the keynote on May 31, ahead of Computex and GTC Taipei in early June, and the stock closed up 6.3% at $224.10 the next day. He introduced RTX Spark, Nvidia's first processor for personal computers, developed with MediaTek on the custom CPU design and with Microsoft on Windows. Laptops and desktops are due in the fall from Asus, Dell, HP, Lenovo, Microsoft Surface and MSI, with Acer and Gigabyte to follow.
This is the first completely re-engineered, reinvented line of PCs that has happened in 40 years.
Alongside the PC push, Huang announced the standalone Vera central processor for data-center AI work, aimed at what he put at a $200 billion market. Nvidia listed NYSE, OpenAI, SpaceXAI, ByteDance, CoreWeave and Oracle Cloud among the customers exploring or planning to adopt the chip, and said Anthropic was evaluating it.
AWS, Google Cloud, Microsoft and Oracle Cloud are among the first named to deploy Vera Rubin instances, in the second half of 2026.
What the shares cost
The third argument is price. Nvidia trades at about 22 times forward earnings, against roughly 25 times for the Nasdaq-100. Estimates of its share of the market for AI accelerators run from 80% to 88%, which makes its results a barometer for the sector.
The stock closed at $206.64 on Aug. 3, near its 50-day and 200-day moving averages, after a closing peak of $235.47 on May 14 and a summer pullback. Semiconductor stocks together shed more than $1 trillion in value during a late-July selloff, and Nvidia has since recovered part of that ground. The average analyst price target is about $303.
The August 26 earnings test
Everything points toward one date. Nvidia reports fiscal second-quarter results after the close on Aug. 26, and management has guided to about $91 billion in revenue for the quarter that ended July 26. Wall Street consensus runs slightly higher, near $92 billion to $95 billion. Revenue in the same quarter a year earlier was $46.7 billion, so the guidance implies growth of about 95%.
The company will also give guidance for the fiscal third quarter, which investors will read for the pace of Blackwell and Rubin shipments.
The risks investors are weighing
The financing push carries its own hazards. Moody's has warned that spending at this scale is eating into free cash flow. Alphabet reported negative free cash flow of $5.9 billion in the second quarter of 2026, its first as a public company, in a period when it spent $44.9 billion on capital projects. The vendor-style financing has drawn comparisons to arrangements Lucent and Nortel ran during the telecom boom of the late 1990s, though in those cases the manufacturer lent the money directly.
Competition and geopolitics add more. Nvidia faces export controls on sales to China, and recent forecasts have explicitly left Chinese data-center compute out of certain projections. AMD is pressing with its Instinct MI400 series, and several hyperscalers are designing custom in-house chips that could reduce their reliance on Nvidia over time.
Nvidia is also defending a patent suit filed in federal court in Texas in April 2026, which alleges wilful infringement of six patents covering AI training and data-center technology.
History offers a caution too. On one commonly cited set of figures, Nvidia's stock fell 68% during the dot-com crash, 85% during the global financial crisis and 66% during the inflation-driven selloff, while other analyses put its worst drawdown closer to 90%. Each recovery eventually carried the shares to new highs, but the pattern is a reminder that even a dominant franchise has delivered steep losses when sentiment turned against it.
What comes next
For now, Nvidia has traded between about $217 and $225 since the Aug. 10 pullback. The Aug. 26 earnings report, measured against $91 billion in guidance, will show whether the three forces lifting the stock this summer can carry it higher into the fall.
Sources
- CNBC: Nvidia, Wall Street asset managers partner on $500B AI push
- Euronews: What Nvidia's $500 billion Wall Street deal signals about the AI boom
- Vantage Markets: NVDA Holds $217 as Nvidia Lands $500B AI Deal
- Vantage Markets: NVDA Stock Analysis August 2026
- The Motley Fool: 3 Reasons Nvidia Stock Could Keep Soaring Through 2026
- Investing.com: Why is NVIDIA stock surging today?
- Windows Forum: Nvidia AI Data Center Boom: Blackwell, Rubin, and 2026 Demand
- Intellectia.ai: Nvidia Earnings Preview August 2026
- Rolling Out: NVIDIA stock is surging in 2026 and here is why it matters








