Nvidia to Raise $500 Billion for AI Infrastructure with Major Wall Street Banks on Board

Apollo, Blackstone, BlackRock, and Goldman Among the Participants

Investment Funds Flow Back into GPU Purchases

Concerns Rise Over "Circular Financing" in AI Investments

Nvidia is joining forces with major Wall Street financial institutions to raise $500 billion (approximately 710 trillion won) to build artificial intelligence (AI) infrastructure. As massive investments are needed in AI data centers, electricity, and semiconductors, Nvidia is expanding its role from just a semiconductor supplier to becoming a core pillar for funding AI infrastructure construction. However, concerns are mounting over so-called "circular financing," in which Nvidia supports client companies’ fundraising efforts and these funds are subsequently used to purchase Nvidia chips.

According to the Financial Times (FT), the Wall Street Journal (WSJ), and Bloomberg on August 10 (local time), Nvidia is working with major U.S. financial firms—including Apollo Global Management, Blackstone, BlackRock, Brookfield Asset Management, Goldman Sachs, and KKR—to pursue a $500 billion fundraising plan aimed at building AI infrastructure.

WSJ, citing sources, reported that Nvidia and these financial players have reached an agreement and could announce the details as early as today. FT also reported that these financial groups plan to form a partnership with Nvidia and invest in AI infrastructure.

It is expected that the $500 billion will not be raised through a single joint fund, but rather that each financial firm will supply capital using different financial instruments. According to WSJ, there is even the possibility that the total amount raised could surpass $500 billion.

As the AI race heats up and the amount of capital needed for data centers, power, and semiconductors surges, tech companies are raising funds not only by issuing stock but also through investment-grade and high-yield corporate bonds, private loans, asset securitizations, and project financing.

Morgan Stanley forecasts that so-called "hyperscalers"—Meta, Microsoft, Alphabet, Amazon—will collectively invest $3.5 trillion in AI infrastructure between 2026 and 2028. Jim Zelter, president of Apollo, recently estimated that more than $8 trillion in capital will be needed for future AI infrastructure construction.

Nvidia Sells Chips and Provides Financing... Concerns Over Circular Financing

Nvidia itself is directly stepping in to provide financial support for its clients’ AI infrastructure investments. Nvidia has been discussing the possibility of providing up to $250 billion in guarantees to allow OpenAI to lease computing capacity at a 10GW data center being built in Ohio, United States.

This kind of structure could become a new growth driver for Nvidia. If Nvidia supports its customers with funding for data center construction and computing, those clients then use the secured financing to purchase Nvidia GPUs, leading to increased revenue for Nvidia.

However, there are warnings that such transactions could artificially inflate demand in the AI market and company valuations. Because companies that Nvidia invests in or helps finance are making large-scale purchases of Nvidia products, funds are circulating within the AI ecosystem itself.

Bloomberg noted that some investors are concerned Nvidia’s circular transactions may be inflating both demand and valuations in the AI industry. Nvidia recently expanded its partnership with SK Group, announcing that the two sides will move forward together on projects exceeding $500 billion in scale.

In the bond market, concerns about Nvidia’s credit risk have also increased. According to Solve, a bond pricing information provider cited by WSJ, the spread between Nvidia corporate bonds and U.S. Treasury bonds of the same maturity has doubled since June to approximately 0.40 percentage points.

If the planned $500 billion fundraising materializes, Nvidia is expected to further expand its influence as the core axis of the AI infrastructure ecosystem—leveraging its dominance in the AI semiconductor market to connect data center construction, power sourcing, and even finance.

This content was produced with the assistance of AI translation services.