AI Compute Demand Drives a New Financing Race as Nvidia, Nebius and CoreWeave Scale Infrastructure

AI Infrastructure Is Moving From a Chip Boom to a Capital-Intensive Compute Buildout

The artificial intelligence expansion is increasingly becoming a financing story as well as a technology story. Nvidia (NVDA) is working with some of Wall Street’s largest financial firms on a framework intended to support as much as $500 billion in AI computing deals, potentially expanding access to the GPUs and infrastructure required by AI developers.

At the same time, earnings from Nebius Group (NBIS), CoreWeave (CRWV) and Super Micro Computer (SMCI) showed strong demand across the AI infrastructure chain. Nebius revenue increased 454% year over year, CoreWeave revenue more than doubled, and Supermicro issued a quarterly sales outlook well above Wall Street expectations. Together, the developments highlight both the scale of AI compute demand and the enormous amounts of capital required to meet it.

Key Points

• Nvidia is working with Goldman Sachs, Blackstone, Apollo, KKR, BlackRock and Brookfield on a framework aimed at financing up to $500 billion in AI computing deals, although no deals had been signed when the initiative was announced.
• Nebius Q2 revenue surged 454% to $582.3 million, while CoreWeave revenue climbed 112% to roughly $2.5 billion and its backlog reached $104.2 billion, highlighting continued demand for AI compute.
• Meeting that demand requires enormous capital: Nebius spent $5.7 billion on capital expenditures during the quarter, while CoreWeave spent $9.4 billion, putting financing and cash requirements alongside growth as key issues for the AI infrastructure buildout.

Nvidia Pushes AI Infrastructure Into the Financing Market

The AI infrastructure expansion has created an increasingly important question: who will finance the enormous amount of computing capacity being planned?

Nvidia is attempting to address that problem by working with some of the largest firms in global finance.

Goldman Sachs (GS), Blackstone (BX), Apollo Global Management (APO), KKR (KKR), BlackRock (BLK) and Brookfield are participating in an initiative intended to facilitate AI computing deals totaling as much as $500 billion.

The framework is particularly relevant to AI startups such as OpenAI and Anthropic, which are important sources of future Nvidia demand as the chipmaker seeks to broaden its customer base beyond major hyperscalers.

The $500 billion figure does not represent signed financing commitments. According to the supplied information, the amount combines deals already under discussion with forecasts for future demand, has no defined time frame, and no transactions had been signed when the initiative was announced.

Individual lenders will also be able to evaluate customers for creditworthiness before committing capital.

Nvidia itself could provide guarantees covering as much as 25% of an opportunity, with projects evaluated individually.

Much of the financing is expected to come from private credit, but the potential scale means public debt markets could also be required. One possible structure involves special-purpose vehicles issuing bonds, potentially worth tens of billions of dollars each, and then leasing Nvidia chips to customers.

The chips and customer offtake agreements could serve as collateral.

That structure could also provide another layer of protection if individual customers cannot meet their obligations because GPUs could potentially be rented to other users.

The initiative effectively adds financial capacity to the AI supply chain. Nvidia supplies the computing hardware, developers need access to it, data centers house it, and financial institutions provide another potential route for funding the infrastructure connecting those pieces.

Nvidia is not alone in pursuing this model.

Broadcom (AVGO) recently announced a financing initiative with Apollo and Blackstone targeting more than 20 gigawatts of compute capacity for frontier AI laboratories through 2028. Broadcom already had $35 billion in financing in place when that partnership was unveiled.

The size of the proposed financing frameworks illustrates how AI infrastructure has expanded beyond semiconductor capital expenditures into private credit, bonds, real estate and other forms of institutional financing.

Why Are Nebius and CoreWeave Growing So Quickly?

The demand behind that financing push can be seen directly in the latest results from AI-focused cloud infrastructure providers.

Nebius reported Q2 revenue of $582.3 million, compared with $105.1 million a year earlier, representing growth of 454%.

Its AI Cloud business generated $574.9 million, or 98.7% of total revenue. Adjusted EBITDA for the AI Cloud segment increased to $285.7 million from $9.5 million.

Nebius also signed four major AI cloud agreements during the quarter with an average total contract value exceeding $1 billion.

Management said the company could potentially sell all of its 2027 capacity but is deliberately preserving some capacity to meet immediate customer requirements.

That level of demand requires heavy investment.

Nebius spent $5.7 billion on capital expenditures during the quarter, driven primarily by GPU purchases and related hardware. The company is building data centers filled with high-performance GPUs and renting that computing capacity to customers.

CoreWeave showed a similar demand pattern at a larger revenue scale.

Quarterly revenue reached roughly $2.5 billion, up from $1.2 billion a year earlier, representing growth of approximately 112%. Adjusted EBITDA reached $1.51 billion with a 59% margin.

Its revenue backlog, including contracted obligations that have not yet been recognized as revenue, increased 246% to $104.2 billion.

CoreWeave also said it signed another $25 billion in customer commitments that will enter its backlog during the third quarter.

Physical capacity is expanding alongside those contracts. Active power increased by nearly 500 megawatts to 1.5 gigawatts, while contracted power stood at roughly 3.7 gigawatts.

But CoreWeave’s expansion carries substantial financial requirements.

Capital expenditures reached $9.4 billion during the quarter, adjusted net losses widened to $567 million from $130 million a year earlier, and quarterly free cash flow was negative $5.74 billion. The supplied material also reports $72.05 billion in liabilities and $640 million in quarterly interest expense.

The numbers demonstrate the central tension in the neocloud model: demand and contracted revenue can grow rapidly, but building the computing infrastructure needed to deliver that capacity requires enormous upfront spending.

What Comes Next for the AI Infrastructure Buildout?

The latest company news suggests that the next phase of AI infrastructure will be shaped not only by demand for GPUs but by the industry's ability to finance and deploy them.

Supermicro provided another signal from a different part of the infrastructure chain.

The server company reported adjusted earnings of $1.70 per share, ahead of expectations of $1.59. Revenue of $11.1 billion came slightly below the projected $11.2 billion.

Its outlook, however, was considerably higher than Wall Street expectations.

Supermicro forecast net sales between $14.5 billion and $15.5 billion, compared with expectations of $11.9 billion.

Demand for the latest generation of computing hardware is also moving forward.

CoreWeave became the first company to validate Nvidia's Vera Rubin NVL72 architecture, while both CoreWeave and Nebius are positioned to deploy Nvidia's next-generation technology through their relationships with the chipmaker.

The supplied material also indicates that older Nvidia hardware may remain economically useful longer than some skeptics have argued. CoreWeave said Nvidia A100 GPUs released in 2020 now have a useful life of at least nine years.

Longer hardware utilization could matter for infrastructure economics because depreciation — the accounting recognition of equipment losing value over time — is an important expense for businesses spending billions of dollars on GPUs.

The broader constraint, however, remains compute availability.

Tech strategist Dan Ives estimated demand relative to chip supply at approximately 13-to-1 or 14-to-1, arguing that enterprises increasingly need to secure their position in line for GPUs and data center capacity.

He also described recent earnings from hyperscalers and neocloud providers as evidence that AI is moving from a capital expenditure phase toward monetization, with AI use cases spreading across industries.

Microsoft (MSFT) was highlighted as one example of enterprises increasing AI investment. Ives estimated that every dollar of capital expenditure can generate a five- to six-dollar revenue multiplier across other parts of the technology industry.

At the same time, funding that expansion remains an unresolved challenge. Companies are still balancing equity and debt as they determine how to pay for increasingly expensive AI infrastructure.

That makes the combination of computing demand, physical capacity, financing availability and eventual revenue generation central to what happens next.

What It Means for Investors

The latest stock market news around AI infrastructure highlights an important evolution in the sector.

The original AI investment cycle centered heavily on demand for Nvidia GPUs. The supplied information shows that the ecosystem surrounding those processors is becoming considerably broader.

Nebius and CoreWeave represent the specialized cloud infrastructure layer, buying GPUs and building data centers designed for intensive AI workloads. Supermicro supplies server infrastructure. Major hyperscalers continue investing in AI capacity. Meanwhile, Goldman Sachs, Blackstone, Apollo, KKR, BlackRock and Brookfield are exploring how institutional capital can finance another wave of computing infrastructure.

The earnings numbers indicate that demand remains strong.

Nebius grew quarterly revenue 454%, CoreWeave grew approximately 112%, and CoreWeave's backlog reached $104.2 billion before another $25 billion of commitments expected to enter the backlog in Q3.

But those figures sit beside equally significant spending requirements.

Nebius invested $5.7 billion in capital expenditures during one quarter. CoreWeave spent $9.4 billion and generated negative free cash flow of $5.74 billion.

Those numbers help explain why Nvidia's proposed financing framework matters.

If AI companies require increasingly large amounts of computing capacity, selling more chips depends not only on customer demand but also on whether customers can finance the infrastructure required to deploy those chips.

That dynamic also explains investor concerns about circular financing. Nvidia has invested in customers including CoreWeave, while the proposed financing initiative could see Nvidia guarantee portions of financing used to acquire AI computing equipment.

The structure could expand the available pool of capital, but the initiative remains preliminary. The $500 billion figure has no fixed time frame, participating lenders can evaluate projects individually, and no deals had been signed when it was announced.

That distinction is important when interpreting the scale of the announcement.

The next major company-specific test identified in the supplied material comes Aug. 26, when Nvidia reports second-quarter earnings.

Conclusion

The AI infrastructure story is expanding from semiconductors into data centers, specialized cloud providers and global capital markets.

Nebius and CoreWeave are demonstrating how quickly demand for purpose-built AI computing can translate into revenue and contractual commitments. Supermicro’s outlook provides another signal of demand further along the hardware infrastructure chain.

But the scale of the required investment is equally clear.

Billions of dollars in quarterly capital expenditures at individual neocloud providers and Nvidia's effort to assemble a financing framework involving some of the world's largest investment firms show that access to capital is becoming increasingly connected to access to compute.

The proposed $500 billion Nvidia initiative remains a framework rather than a pool of committed capital. Yet its existence illustrates the size of the infrastructure challenge confronting the industry.

For the AI market, the emerging question is no longer simply how many GPUs companies want. It is also how quickly the industry can finance, build and monetize the computing infrastructure required to put those GPUs to work.

FAQs

What is Nvidia’s $500 billion AI financing initiative?

Nvidia is working with Goldman Sachs, Blackstone, Apollo, KKR, BlackRock and Brookfield on a framework intended to facilitate as much as $500 billion in AI computing deals. The figure combines potential deals and forecasts of future demand, has no fixed time frame, and no deals had been signed when the initiative was announced.

How fast is Nebius growing?

Nebius reported Q2 revenue of $582.3 million, up 454% from $105.1 million a year earlier. Its AI Cloud business generated $574.9 million, while the company signed four major AI cloud agreements with an average total contract value exceeding $1 billion.

How large is CoreWeave’s backlog?

CoreWeave reported a revenue backlog of $104.2 billion, up 246% year over year. The company also said it had signed another $25 billion in customer commitments that are expected to enter its backlog during the third quarter.

Why are AI infrastructure companies spending so much money?

AI infrastructure providers need large quantities of GPUs and related data center hardware to expand computing capacity. Nebius spent $5.7 billion on capital expenditures during the quarter, primarily on GPUs and related hardware, while CoreWeave spent $9.4 billion.

When does Nvidia report its next earnings?

According to the supplied material, Nvidia is scheduled to report its second-quarter earnings on Aug. 26.

This article was created with AI assistance and reviewed by an editor. For details, please refer to our Terms of Use.

Go Beyond the Market Brief with Market Edge

Follow SharperTrades’ complete approach to trading and investing, combining active trade opportunities through Block Orders, long-term research through Stock Investor, and structured market education through the Swing Trading Masterclass. Try Market Edge for $19 your first month →

Explore Research with Stock Investor

Stock Investor is SharperTrades’ platform for long-term investing research and portfolio management. Members receive research reports, portfolio updates, conviction tracking, and in-depth analysis designed to support disciplined investment decisions.

Explore Active Trading & Income Strategies

Block Orders tracks institutional activity and highlights active trade setups and price behavior across long and short opportunities.

For options-focused traders, Essential Option Income provides a structured approach to options income strategies, while Pro Option Trader offers a broader range of options strategies and trade opportunities.

Think More Clearly with SteadyCapital

SteadyCapital is SharperTrades’ decision-support system for long-term investors, built around the SteadyCapital Method™. Review investment ideas, challenge assumptions, evaluate valuation and risk, compare companies, and think through important buy, hold, add, trim, or sell decisions before you act.

Risk Disclosure

All content is provided for educational purposes only and does not constitute investment advice. Trading involves risk, and past performance is not indicative of future results. Please review our full Risk Disclosure for additional information.