Huang Renxun Responds to "US$500 Billion Financing" Doubt: Genuine Demand Exists, AI Factory is Becoming an Investable Infrastructure Asset

BlockBeats News, August 11, Nvidia CEO Jensen Huang stated that Nvidia has partnered with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR to establish an independent financing platform. The plan is to mobilize over $500 billion in third-party capital in the long term to support AI infrastructure development. Huang emphasized that this number is not Nvidia's revenue, the size of a single fund, or a financing commitment for a single customer.

Huang said that AI is transitioning from the research phase to large-scale production, and AI factories are becoming investable infrastructure assets. The Nvidia AI factory consists of accelerated computing, networking, system software, AI frameworks, and the developer ecosystem, serving different customers and workloads and allowing redeployment among customers, cloud service providers, and operators. He believes that in the AI field, "computing power is revenue."

Regarding market concerns about "cyclical financing," Huang stated that the design purpose of this plan is to introduce independent, long-term institutional capital. Financial institutions will independently evaluate each project's customer, actual needs, equipment utilization, cash flow, and residual value and make independent decisions on whether to provide funding. Nvidia is responsible for providing the AI factory platform, not for determining how capital is allocated. He said the demand comes from cutting-edge AI labs, AI-native companies, cloud service providers, traditional enterprises, and countries building AI services.

At the same time, Huang acknowledged that in some projects, Nvidia may provide up to a maximum of 25% of the project's size as "residual value support." However, the related support will be carefully evaluated for each project and will only be used to complement, not replace, financial institutions' independent underwriting. He stated that the reason for being able to provide such support is that Nvidia's computing power is versatile, globally adopted, upgradable through software updates, and can be redeployed among different customers, thereby helping to maintain the residual value of the equipment.

Regarding whether the market can absorb such a large-scale addition of computing power, Huang said the key is not "whether to build data centers" but whether these facilities can become AI factories with real output. Financing partners will still evaluate around actual customer needs, utilization rates, cash flow, and residual value. The added capacity will be built on actual customer economic models. He believes that AI has begun to create real value in areas such as software development, drug discovery, product design, customer service, and automated operations, stating that "more computing power leads to better AI, better AI leads to more usage, more usage leads to more revenue."