Amazon pursues $8 billion sale-and-leaseback for Nvidia AI chips

  • Amazon is pursuing a proposed sale-and-leaseback for approximately $8 billion of Nvidia AI chips.
  • The transaction would transfer the chips to outside investors while Amazon continues leasing and using them.
  • Morgan Stanley reinstated Nvidia as a Top Pick, citing AI data-center demand and capacity expansion.

Amazon is pursuing a proposed sale-and-leaseback involving approximately $8 billion, or roughly 10.7 trillion won, of Nvidia artificial-intelligence chips. Outside investors would acquire the hardware while Amazon continues to lease and use it, potentially making the cloud provider’s balance sheet more asset-light as AWS expands AI capacity. The newer account describes the transaction as a financing vehicle, while the earlier account had characterized it as an established $8 billion special purpose vehicle that would buy Grace Blackwell chips and lease them back to AWS clients. The structure reflects broader use of debt, private capital and asset-backed financing as cloud companies and hyperscalers fund increasingly capital-intensive AI data centers. Morgan Stanley has reinstated Nvidia as a Top Pick, citing AI data-center demand and capacity expansion. Amazon also signed a 20-year nuclear power agreement with Constellation Energy and a multiyear agreement worth more than $1 billion with Synopsys, supporting its broader AI infrastructure and custom-silicon strategy. The related analysis presents Nvidia as a data-center-scale AI infrastructure provider and discusses its US$500 billion ecosystem financing plans, a $328 fair value, a separate cash-generation valuation and risks involving vendor financing, data-center constraints and customer vertical integration. Simply Wall St says its analysis relies on historical data and analyst forecasts, is not financial advice or a recommendation to trade Nvidia shares, and may not include the latest price-sensitive announcements or qualitative information.

The information on this website is generated using AI and we cannot guarantee its accuracy. Please use it as reference information only.