Amazon Explores $8 Billion Nvidia AI Chip Deal

Amazon is seeking to transfer about $8 billion worth of advanced Nvidia Artificial Intelligence chips to investors through a new financing vehicle, the Financial Times reported, as the cloud giant looks for ways to strengthen its balance sheet while continuing to expand its AI Infrastructure.

According to the report, Amazon has recently held discussions with investors to assess interest in a transaction that would move thousands of Nvidia Grace Blackwell chips into a special-purpose vehicle, or SPV.

Amazon Explores $8 Billion Nvidia Chip Deal

The proposed transaction would involve Nvidia’s advanced Grace Blackwell chips that Amazon is installing across its US data-centre network.

Rather than continuing to hold the expensive semiconductor assets directly on its balance sheet, Amazon could transfer them to an SPV funded partly through outside investment and debt issuance.

The structure would allow Amazon to continue using the chips while investors would provide financing for the assets.

The Financial Times reported that Amazon is considering the transaction as part of a broader effort to adopt a more asset-light approach to its balance sheet.

How the Proposed Financing Structure Could Work

Under the reported structure, the Nvidia chips would be transferred to a special-purpose vehicle established specifically for the transaction.

The SPV would seek financing from outside investors, including through debt issuance. Amazon would then lease the chips back from the vehicle so that it could continue using them in its data centres.

This type of financing structure can allow a company to unlock capital tied up in expensive physical assets while retaining operational access to those assets.

Amazon is reportedly considering offering investors an equity stake of up to 10% in the vehicle.

Thousands of Nvidia Grace Blackwell Chips Involved

The chips involved in the proposed transaction are Nvidia Grace Blackwell systems, which are designed for demanding artificial intelligence and accelerated computing workloads.

According to the Financial Times report, Amazon has bought or leased the chips and installed them across more than a dozen US data centres in five states.

The facilities include data centres in Nevada and Virginia, highlighting the scale of Amazon’s ongoing investment in AI computing infrastructure.

Moving those assets into an investment vehicle would not mean Amazon would stop using them. Instead, the company would reportedly lease them back from the SPV after the transaction.

Why Amazon Is Considering an Asset-Light Approach

Artificial intelligence infrastructure requires enormous amounts of capital. Advanced AI chips, data centres, networking equipment and power infrastructure can require substantial upfront investment before they generate revenue.

For cloud companies such as Amazon, the rapid growth of AI workloads has increased the need for computing capacity. At the same time, large infrastructure investments can put pressure on cash flow and balance-sheet resources.

An asset-backed financing structure can potentially allow Amazon to continue expanding its computing capacity while transferring ownership of some expensive equipment to outside investors.

The approach could therefore provide Amazon with additional financial flexibility without requiring it to stop using the underlying AI hardware.

Nvidia Chips Are at the Centre of the AI Infrastructure Race

Nvidia’s advanced processors have become central to the infrastructure used to train and run sophisticated AI models.

Companies developing generative AI systems and cloud providers expanding AI services have been investing heavily in Nvidia hardware because advanced GPUs and related systems provide the computing capacity required for large-scale AI workloads.

Amazon has been expanding its own AI infrastructure as demand for cloud computing and AI services increases. The proposed financing arrangement reflects the capital-intensive nature of that expansion.

Amazon Could Retain Access to the Chips

A key feature of the reported structure is that Amazon would continue using the chips after transferring them to the special-purpose vehicle.

The cloud company would lease the hardware back from the SPV, allowing its data centres to continue operating with the same computing infrastructure.

For investors, the transaction would create exposure to physical AI infrastructure through the vehicle, while Amazon would effectively pay for continued use of the equipment through lease arrangements.

The exact financial terms, including the lease costs, debt structure and potential investor returns, have not been disclosed.

Amazon May Offer Up to 10% Equity Stake

The Financial Times reported that Amazon plans to offer investors an equity stake of up to 10% in the vehicle.

The remaining financing could come through debt raised by the SPV, allowing the vehicle to acquire or hold the Nvidia chips without requiring Amazon to finance the entire asset base itself.

The final ownership structure could change because discussions with potential investors are still ongoing.

What the Deal Could Mean for Amazon’s Balance Sheet

The proposed transaction could change how Amazon accounts for and finances part of its rapidly expanding AI infrastructure.

By transferring expensive semiconductor assets to an investment vehicle, Amazon could potentially reduce the amount of capital directly tied up in those assets.

The strategy could also provide funds that Amazon can deploy toward other business priorities, including additional cloud infrastructure, AI services and technology investments.

However, an asset transfer accompanied by a leaseback would also create ongoing lease obligations. The financial impact would therefore depend on the final terms of the transaction.

Amazon’s AI Infrastructure Spending

Amazon has been investing heavily in data centres and computing infrastructure to support the growth of its cloud business and artificial intelligence services.

AI workloads require specialized processors and large amounts of computing capacity. Demand for these resources has increased as businesses adopt generative AI applications and developers build increasingly sophisticated AI models.

The spending requirements have created a broader financing challenge for major technology companies. Businesses need to secure access to large quantities of advanced chips while balancing infrastructure investment against cash generation and shareholder returns.

Investors’ Role in AI Infrastructure Financing

The reported Amazon transaction reflects a broader potential role for institutional investors in financing AI infrastructure.

Instead of technology companies funding every piece of infrastructure directly, specialized investment vehicles can potentially provide capital for assets such as servers, chips and data-centre equipment.

Investors may be attracted to such structures if the assets generate predictable lease payments and retain value over their useful life.

At the same time, investors would need to assess risks related to technological obsolescence, chip values, lease payments and the rapidly changing AI hardware market.

Amazon and Nvidia Yet to Comment

Amazon and Nvidia did not immediately respond to Reuters’ requests for comment outside regular business hours.

The Financial Times report was based on people familiar with the matter. Discussions with investors are reportedly still at an exploratory stage, meaning the proposed transaction could change or may not ultimately be completed.

What to Watch Next

The next key developments will be whether Amazon reaches agreements with investors and whether the proposed special-purpose vehicle moves forward.

Investors will also be watching the final value of the assets transferred, the amount of debt raised by the SPV, Amazon’s equity contribution and the terms under which Amazon leases the chips back.

The transaction could provide an example of how major technology companies are seeking new ways to finance the enormous infrastructure requirements created by the artificial intelligence boom.

Frequently Asked Questions

1. How much Nvidia chip infrastructure is Amazon reportedly seeking to offload?

Amazon is reportedly seeking to transfer about $8 billion worth of advanced Nvidia chips to investors through a new special-purpose vehicle.

2. Which Nvidia chips are involved?

The proposed transaction involves Nvidia Grace Blackwell chips that Amazon is installing in US data centres.

3. What is a special-purpose vehicle?

A special-purpose vehicle, or SPV, is a separate legal entity created for a specific financial transaction or group of assets. It can be used to raise financing and hold assets separately from the main company.

4. Will Amazon stop using the Nvidia chips?

No. Under the reported structure, Amazon would lease the chips back from the SPV and continue using them in its data centres.

5. Why is Amazon considering the transaction?

The reported objective is to strengthen Amazon’s balance sheet and adopt a more asset-light approach while continuing to use expensive AI infrastructure.

6. How much equity could Amazon offer investors?

According to the Financial Times report, Amazon plans to offer investors an equity stake of up to 10% in the vehicle.

7. Where are the chips installed?

The chips are reportedly installed in more than a dozen Amazon data centres across five US states, including Nevada and Virginia.

8. Has Amazon confirmed the reported deal?

No. The report describes discussions with investors, and Amazon and Nvidia had not immediately responded to Reuters’ requests for comment outside regular business hours. The structure and terms could therefore change.