Apollo Global Management is in advanced discussions to secure a $3.4 billion loan that would finance the purchase of Nvidia artificial intelligence chips, which would then be leased to Elon Musk's xAI, according to a report from The Information. The deal, which could be finalized this week, highlights the growing trend of specialized financing for AI hardware as demand for computing power surges.
In a parallel development, Alphabet is reportedly weighing a $15 billion high-grade bond offering in the United States. This move is part of a broader pattern where major cloud providers are tapping debt markets to fund massive investments in AI infrastructure. Hyperscale cloud companies are projected to spend over $630 billion on AI this year, even as the financial returns from these expenditures have yet to fully materialize.
Alphabet recently set a aggressive tone, forecasting 2026 capital expenditures between $175 billion and $185 billion, a significant increase from $91.45 billion in 2025. CEO Sundar Pichai emphasized that these AI investments are driving revenue growth across the company, noting a 48% jump in Google Cloud revenue to $17.7 billion for the December quarter.
While Nvidia remains the primary supplier for large-scale AI training workloads, some investors are betting on increased competition. Analysts point to the rise of custom application-specific integrated circuits (ASICs), like Google's Tensor Processing Units (TPUs), designed for specific AI tasks such as inference. Broadcom has positioned itself as a key enabler of this trend, having disclosed a $10 billion AI chip order from an unnamed customer last September, with analysts speculating it could be OpenAI.
The massive capital outlays are causing some investor unease, reviving concerns that cloud giants are transitioning from asset-light to capital-intensive business models. Amazon has outlined approximately $200 billion in planned capital expenditure for the year. Despite these worries, chip stocks, including Nvidia, AMD, and Broadcom, rallied recently as markets bet sustained infrastructure spending will continue to fuel demand.
The key question for the market is whether the flood of capital into AI chips and data centers will translate into lasting cloud revenue growth and profitability, or if it will result in an oversupply of expensive, underutilized hardware.
