Amazon Looks To Move $8 Billion In AI Chips Off Its Balance Sheet

Amazon is exploring an unconventional financing arrangement that would transfer roughly $8 billion worth of advanced Nvidia artificial-intelligence chips to outside investors while allowing the company to continue using the hardware in its U.S. data centers. Under the proposed structure, thousands of Nvidia Grace Blackwell chips would be transferred to a special-purpose vehicle financed largely through debt, with Amazon leasing the processors back rather than continuing to own them directly. The proposal comes as Amazon undertakes an extraordinary AI infrastructure expansion, with capital expenditures expected to reach roughly $200 billion to $220 billion in 2026, much of it directed toward AWS, chips and data centers. Amazon has already raised tens of billions of dollars through bonds and secured a $17.5 billion delayed-draw loan facility this year. The prospective chip transaction illustrates how the enormous cost of the AI race is changing Big Tech finance, pushing even cash-rich companies toward debt, leases and other financing structures designed to preserve balance-sheet flexibility while continuing the massive infrastructure buildout.

Key Takeaways

Amazon is considering transferring approximately $8 billion of Nvidia Grace Blackwell AI chips housed across more than a dozen U.S. data centers into a special-purpose vehicle funded by outside investors, then leasing the processors back.
The transaction would shift some ownership, financing and depreciation risk away from Amazon’s balance sheet as the company commits roughly $200 billion or more to 2026 capital expenditures, largely associated with AWS and AI infrastructure.
The proposal reflects a broader transformation in Big Tech financing, as companies increasingly supplement their enormous cash flows with bonds, loans, leases and specialized financing vehicles to fund the unprecedented capital requirements of the AI boom.

In-Depth

Amazon’s reported plan to move roughly $8 billion of Nvidia chips into an outside investment vehicle underscores just how capital-intensive the artificial-intelligence race has become. Even one of the world’s largest and most profitable technology companies is searching for ways to finance AI infrastructure without carrying every expensive asset directly on its balance sheet.

The proposed arrangement would place thousands of Nvidia Grace Blackwell processors into a special-purpose vehicle financed by outside investors. Amazon would continue using the chips by leasing them back. Such a structure could preserve financial flexibility while transferring some ownership and depreciation risk to investors.

The larger numbers are staggering. Amazon expects roughly $200 billion or more in 2026 capital expenditures, with much of that investment supporting AWS, AI chips and data-center capacity. The company has already tapped conventional financing aggressively, including a $17.5 billion delayed-draw loan facility and major bond offerings.

There is nothing inherently troubling about sophisticated financing. Private capital exists precisely to allocate resources toward opportunities investors believe can produce adequate returns. But the scale of AI investment should encourage discipline rather than complacency. Chips depreciate quickly, technological generations change rapidly, and today’s premium hardware can become comparatively outdated within several years.

Amazon argues that substantial customer commitments support its enormous infrastructure investment. That matters. Still, moving expensive hardware into specialized financing vehicles demonstrates that the AI boom is no longer simply a technology story. It is increasingly a credit-market story as well, testing whether future AI revenues can justify the extraordinary amount of capital now being committed.

Subscribe to Updates

What's Hot

Amazon Looks To Move $8 Billion In AI Chips Off Its Balance Sheet

Key Takeaways

In-Depth

Sources

Related Posts