SpaceX in talks to raise $40 Billion debt to Buy Nvidia Chips

Source and verification note (7 October 2026): The proposed $40 billion financing is based on the Financial Times’ October 6 original report, republished by Yahoo Finance, which cites unnamed people familiar with talks. Reuters explicitly attributes the same financing details to the FT, so it is not independent confirmation. SpaceX, Apollo, Nvidia and Pimco have not announced the arrangement; its amount, participants and timing could change, and no agreement or chip order is confirmed here. Separate historical figures in this article come from SpaceX’s Q2 results filed with the SEC and its June bond announcement; those documents do not verify the new financing talks.

SpaceX is reportedly seeking around $40 billion in financing to purchase Nvidia AI chips, in what could become one of the largest debt financings tied to the global artificial-intelligence infrastructure buildout. The financing effort is being led by Apollo Global Management and is expected to combine approximately $10 billion of bank loans with $30 billion of investment-grade debt.

The proposed financing is still in an early stage and could ultimately change or fail to close, according to people familiar with the discussions. If completed, however, the deal would demonstrate the enormous amount of capital SpaceX is willing to deploy on computing infrastructure as Elon Musk expands the company’s business beyond rockets and satellite connectivity into AI computing, data centers and AI services.

Key takeaways

• SpaceX is reportedly seeking about $40 billion to finance Nvidia AI-chip purchases.
• The proposed package includes roughly $10 billion of bank loans and $30 billion of investment-grade debt.
• Apollo Global Management is expected to lead the financing.
• PIMCO is among the investors reportedly evaluating the debt.
• The transaction is expected to close in 2027 if completed.
• SpaceX has said it plans to build its AI data centers exclusively with Nvidia hardware.
• AI capital expenditure reached $15.83 billion in Q2 2026, according to SpaceX’s reported results.
• SpaceX’s total capital expenditure reached $28.48 billion during the first half of 2026.
• The company generated $2.56 billion in AI revenue during Q2, while the AI segment recorded a $1.26 billion operating loss.
• SpaceX already raised $25 billion through its inaugural bond offering in June.
• The proposed financing highlights the growing use of debt and structured capital to fund AI infrastructure.

SpaceX wants another $40 billion for Nvidia chips

The reported financing would be specifically tied to SpaceX’s rapidly expanding AI-computing ambitions.

According to the Financial Times, SpaceX is seeking approximately $40 billion through a financing package led by Apollo Global Management. The proposed structure consists of about $10 billion in bank loans and another $30 billion of investment-grade debt.

PIMCO is among the financial institutions reportedly involved in discussions around the financing.

The transaction is expected to close in 2027, although negotiations remain at an early stage. SpaceX, Nvidia, Apollo and PIMCO had not immediately confirmed the reported financing.

That distinction is important. The $40 billion figure represents a reported financing target rather than a completed borrowing transaction.

If the financing proceeds, the money would help SpaceX purchase Nvidia’s advanced AI processors as it builds increasingly large computing clusters.

Why SpaceX needs so many Nvidia chips

SpaceX is no longer investing in AI as a small side business.

The company’s second-quarter results showed just how rapidly AI infrastructure has become a major part of its capital allocation.

SpaceX reported $18.37 billion in total capital expenditure during the second quarter of 2026. Of that amount, $15.83 billion went toward its AI segment.

That means AI represented roughly 86% of SpaceX’s quarterly capital expenditure.

For the first six months of 2026, SpaceX reported total capital expenditure of $28.48 billion, including $23.55 billion for AI.

The figures represent a dramatic increase from the previous year. AI capital expenditure was only $749 million in the second quarter of 2025.

The company’s AI investment is therefore expanding at extraordinary speed.

SpaceX’s AI business is already generating revenue

The enormous spending is not being made without a revenue opportunity.

SpaceX reported $2.56 billion in AI revenue during the second quarter, up sharply from $737 million a year earlier.

AI solutions and infrastructure generated $2.19 billion of the quarterly total, while advertising contributed $367 million.

The company also reported that its AI business had $14.1 billion of contracted sales through several cloud-services agreements.

These contracts are important because SpaceX is increasingly positioning computing capacity as a commercial product.

Instead of building data centers solely for internal AI workloads, the company can sell computing capacity to outside customers.

That creates a potential business model similar to cloud infrastructure providers: spend heavily upfront on chips and data centers, then recover that investment by selling computing capacity over time.

However, the economics are still developing.

SpaceX’s AI segment recorded a $1.26 billion operating loss in the second quarter, despite generating $2.56 billion in revenue.

The company did report positive AI adjusted EBITDA of $1.15 billion, but the gap between adjusted EBITDA and operating income reflects the enormous depreciation, research and development and other costs associated with the infrastructure buildout.

Musk has committed SpaceX to Nvidia hardware

The proposed financing also follows a clear hardware strategy from Musk.

During SpaceX’s August earnings call, Musk said the company had decided to build its AI infrastructure exclusively around Nvidia technology because he considered Nvidia’s Vera Rubin architecture the best available AI computing architecture.

That commitment makes the reported $40 billion chip financing more understandable.

Building large AI data centers requires enormous numbers of high-performance accelerators. Nvidia’s processors are among the most sought-after components for training and running advanced AI models.

For Nvidia, a buyer willing to commit tens of billions of dollars to its hardware represents an unusually large source of demand.

For SpaceX, however, the decision also creates concentration risk.

If the company relies overwhelmingly on one chip supplier, its AI expansion becomes exposed to Nvidia’s product availability, pricing, delivery schedules and technological roadmap.

Colossus is at the center of the expansion

A major component of Musk’s AI strategy is the Colossus computing infrastructure associated with xAI.

SpaceX’s second-quarter results said its AI computing capacity had reached 1.4 gigawatts, compared with 1.0 gigawatt in the first quarter.

The company has been rapidly expanding Colossus II and other AI infrastructure.

Musk has also indicated that SpaceX expects to have more than 2 gigawatts of computing capacity by the end of 2026 and close to 10 gigawatts by the end of 2027.

The scale is significant because a gigawatt of computing capacity represents an enormous concentration of processors, power infrastructure, cooling systems, networking equipment and data-center construction.

The proposed Nvidia financing therefore isn’t simply about buying individual chips.

It is effectively part of a much larger infrastructure expansion involving data centers, electricity and networking systems.

SpaceX is turning computing capacity into a business

One of the most important aspects of the strategy is that SpaceX wants to monetize the infrastructure it builds.

The company has entered into cloud-services agreements that allow customers to access its computing capacity.

Its second-quarter results said new cloud-services agreements generated $1.6 billion in incremental AI infrastructure revenue during the quarter.

This changes the financial logic behind the spending.

A traditional AI company might borrow money to buy computing hardware primarily to train its own models.

SpaceX is attempting to create a broader infrastructure business in which the same computing assets can generate revenue from external customers.

The company can potentially use the infrastructure to support xAI and Grok while also selling spare or dedicated capacity to third parties.

That model could make the massive capital requirements more defensible if utilization and pricing remain high.

The debt burden is becoming a bigger question

The proposed $40 billion financing also raises questions about SpaceX’s balance sheet.

The company already completed a $25 billion inaugural bond offering in June.

According to SpaceX’s official announcement, the bonds consisted of multiple tranches maturing between 2031 and 2056, with interest rates ranging from 5.35% to 6.65%. The proceeds were primarily intended to repay an existing bridge facility, with remaining funds available for general corporate purposes.

SpaceX also received investment-grade ratings from major agencies following its IPO. Moody’s assigned Baa1, Fitch assigned BBB+ and S&P assigned BBB, all with stable outlooks.

Investment-grade status is important because it potentially allows pension funds, insurers and other large institutional investors to purchase SpaceX debt.

However, investment-grade debt is not risk-free.

SpaceX’s aggressive AI capital spending means investors must assess whether future cash generation can justify the scale of investment.

S&P specifically identified the AI business as an area of uncertainty because of its high capital requirements and competitive environment.

SpaceX’s cash generation has to catch up

The central financial question is whether SpaceX can convert its AI infrastructure investment into sufficiently high and predictable cash flows.

The company’s second-quarter results provide both positive and negative signals.

Revenue increased substantially, reaching $7.81 billion for the quarter. AI revenue increased rapidly, and the company said it was securing major cloud-services contracts.

At the same time, capital expenditure reached $18.37 billion in a single quarter.

The scale of investment is therefore significantly ahead of the business’s current cash-generating capacity.

SpaceX management has argued that new AI-compute investments can have payback periods of less than a year.

If that assumption proves correct, the company could potentially recycle cash generated from new computing infrastructure into additional capacity.

If demand grows more slowly, however, SpaceX could find itself carrying large amounts of debt against expensive computing assets whose economic returns take longer to materialize.

The proposed deal fits a larger AI financing trend

SpaceX’s reported financing is part of a much larger shift in the way AI infrastructure is being financed.

The cost of building AI capacity has become too large for many companies to fund entirely from existing cash flow.

That has encouraged banks, private-credit firms, asset managers and institutional investors to develop financing structures specifically for chips and data centers.

Nvidia itself announced a partnership in August with major financial institutions including Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR to develop financing pools for AI infrastructure.

The consortium has discussed creating large pools of capital that could help Nvidia customers finance chips and associated infrastructure.

Apollo’s involvement in the reported SpaceX transaction therefore fits into a wider trend toward financing AI hardware as an infrastructure asset rather than treating it simply as an operating expense.

Apollo already has a relationship with Musk’s AI businesses

Apollo is not a new participant in Musk’s AI expansion.

Business Standard reported that Apollo previously led a $7 billion financing associated with xAI, with the debt used through a special-purpose vehicle to help fund access to Nvidia chips for the Colossus 2 data center.

That experience could make Apollo particularly familiar with the financing structure and risk profile of Musk’s AI infrastructure plans.

It also explains why the asset manager could be positioned to coordinate a much larger financing package for SpaceX.

For Apollo, the transaction could represent another major opportunity to deploy capital into AI infrastructure.

For SpaceX, Apollo’s role could help distribute the borrowing across a wider base of institutional investors.

Nvidia stands to benefit from the spending wave

The proposed financing is also significant for Nvidia.

AI companies and infrastructure operators are competing for access to Nvidia’s latest processors as they expand data-center capacity.

A $40 billion financing package dedicated to Nvidia hardware would represent a major potential order.

SpaceX’s commitment to using Nvidia hardware exclusively for its data centers strengthens that relationship.

Nvidia is simultaneously attempting to broaden the financing ecosystem around its products because the cost of building AI infrastructure is becoming a constraint for customers.

The result is a new financial model in which chip suppliers, banks, private-credit firms and institutional investors increasingly interact around the same AI infrastructure projects.

What could go wrong?

The biggest risk is that AI computing capacity grows faster than profitable demand.

SpaceX is betting that customers will continue paying significant amounts for access to AI compute.

That requires strong demand from AI model developers, cloud providers and enterprise customers.

There is also a technology risk.

AI hardware evolves rapidly. A data center filled with today’s most powerful processors can become less economically attractive if newer chips deliver substantially better performance per dollar and per watt.

That means SpaceX has to keep upgrading infrastructure while recovering the cost of existing hardware.

Financing adds another layer of risk because debt creates fixed obligations regardless of how quickly AI revenue grows.

Finally, SpaceX is simultaneously funding rockets, Starlink expansion and AI infrastructure. The company therefore has multiple capital-intensive businesses competing for resources.

What the $40 billion financing means for SpaceX

The proposed financing is best understood as a bet on AI infrastructure rather than simply a large chip purchase.

SpaceX is attempting to build an integrated business spanning satellites, connectivity, rockets, AI models and computing infrastructure.

Its strategy is to use the company’s existing engineering capabilities and capital access to build infrastructure at enormous scale.

The potential $40 billion financing would provide another major source of capital for that strategy.

But it also means investors will increasingly judge SpaceX not only by rocket launches and Starlink growth, but by whether its AI investments generate attractive returns.

That is a significant change for a company historically defined by space transportation and satellite communications.

The Bigger Picture

SpaceX’s reported $40 billion Nvidia financing highlights how quickly AI has evolved from a software story into a capital-intensive infrastructure industry. The proposed borrowing would be larger than many technology companies’ entire annual capital budgets and would come on top of SpaceX’s $25 billion bond offering earlier in 2026.

The bigger issue is not simply how many Nvidia chips SpaceX buys. It is whether the company can turn those chips into a commercially profitable computing platform quickly enough to justify the debt required to acquire them.

The strategy has potentially powerful economics if AI demand remains strong. SpaceX can build computing capacity, use it internally for xAI and Grok, and sell excess capacity to outside customers. But the same strategy exposes the company to enormous capital requirements, hardware obsolescence and the possibility that AI infrastructure supply eventually grows faster than demand.

Looking Ahead

The next important milestone will be whether SpaceX and its financing partners can turn the reported discussions into a finalized $40 billion transaction. The reported package is still at an early stage, and the final amount, structure, lenders and closing date could change.

If completed, the financing would reinforce SpaceX’s position as one of the world’s largest private buyers of AI computing infrastructure and provide Nvidia with another enormous source of demand. It would also give investors a much clearer test of Musk’s strategy: whether massive debt-funded AI infrastructure spending can translate into the high-growth, high-margin computing business SpaceX is targeting.