Apollo, Blackstone Partner With Nvidia on $500 Billion AI Infrastructure Financing Deal

Apollo Global and Blackstone are partnering with Nvidia on a $500 billion financing package to accelerate AI infrastructure development, the Financial Times reported on August 10, 2026. The deal, among the largest private capital mobilizations in tech history, would fund data centers, chips, and energy systems, reflecting a shift to structured finance for AI compute. Neither company has commented, and terms are still being negotiated.

August 10, 2026 , Inside AI — A consortium of Wall Street heavyweights is teaming up with Nvidia to assemble a massive $500 billion financing package aimed at accelerating artificial intelligence infrastructure development, the Financial Times reported on Monday. Apollo Global and Blackstone are among the financial groups working with Nvidia on the deal, which would rank among the largest private capital mobilizations in tech history. The exact structure remains fluid, but the scale signals a decisive shift in how foundational AI compute capacity gets funded. The funding would target the physical backbone of AI: data centers, specialized chips, and energy systems required to train and run increasingly large models. Nvidia’s chips power most advanced AI workloads, giving the company unique leverage to coordinate such an effort. Neither Nvidia, Apollo, nor Blackstone has publicly commented on the discussions. The Financial Times cited people familiar with the matter, who cautioned that terms are still being negotiated and the final amount could change. Infrastructure Demands Outstrip Traditional Budgets The proposed $500 billion figure dwarfs recent infrastructure investments. Last year, Microsoft and BlackRock launched a $100 billion AI infrastructure partnership. Amazon, Google, and Microsoft each plan to spend over $50 billion annually on data center expansion. Nvidia’s involvement suggests a move beyond chip sales into orchestrating the financial architecture that makes large-scale AI deployment possible. The company’s market capitalization of roughly $3 trillion gives it enormous influence, but direct capital expenditure at this level typically requires outside partners. Apollo and Blackstone together manage over $1.5 trillion in assets, with deep experience in infrastructure and credit. Their participation signals that private equity sees AI hardware as a long-term yield play, not a speculative bubble. What’s Fueling the Sudden Rush Training frontier models like GPT-5 or Gemini Ultra requires clusters of 100,000 GPUs or more. Each cluster can cost over $2 billion in hardware alone, before real estate, power, and cooling. The industry faces a projected 50-gigawatt power deficit for data centers by 2030. This financing model echoes the 1990s telecom buildout, when consortia funded fiber networks that later underpinned the internet. Today’s AI infrastructure carries similar “backbone” characteristics, with high upfront costs and decades-long utility. Wall Street’s embrace also reflects a maturation of AI as an asset class. Rather than relying solely on tech company balance sheets, the industry is turning to structured finance, project bonds, and joint ventures to spread risk and accelerate deployment. Regulatory hurdles remain. Large-scale energy projects face permitting delays, and the concentration of AI compute among a few players has drawn antitrust scrutiny in Washington and Brussels. Any deal of this size would likely trigger CFIUS review if foreign investors are involved. Still, the sheer magnitude of the reported package underscores a growing consensus: AI infrastructure is no longer a niche corporate expense but a foundational layer of the global economy, requiring capital markets to step in where corporate budgets fall short.