In the financial world, reinvestment strategy is essential to maintain long-term dominance. Jensen Huang’s company has perfected a growth model where its own profits feed the ecosystem of startups that, in turn, become its best customers.
Behind its meteoric stock rise is a deliberate investment strategy: investing billions in AI startups that often end up being crucial customers for their GPUs. The result is an ecosystem where capital and demand feed off each other.
NVentures: the investment machine
The corporate venture capital arm, NVentures, acts as the driving force behind that strategy. Its formal mission is to support “technological visionaries” in AI, biotechnology and robotics, but its formula combines capital with preferential access to Nvidia resources (especially its GPUs) to accelerate the growth of startups.
The effect is twofold: it drives technology adoption and roots those companies in the Nvidia ecosystem from their early stage.
On the fly of numbers
As of November 2025, NVentures exhibits a portfolio of some 53 companies and a growing investment pace: more than 50 deals registered in 2025, surpassing 48 in 2024, with 23 new entries in 2025 alone.
Many of these agreements are accompanied by hardware purchase commitments that bind the recipient companies to Nvidia’s demand.
“Circular financing”
Here the controversy arises: Nvidia invests in companies that, in turn, allocate part of those funds to acquire their hardware. Analysts and critics call this “circular financing” or even “self-consumption,” and warn that it can create artificial demand based more on capital flows than on organic adoption.
The examples cited are large in scale: from the $100 billion commitment to OpenAI in September 2025 to deploy data centers powered by Nvidia systems, to participation in multi-million dollar rounds such as those related to xAI.
Other notable beneficiaries mentioned include Mistral AI, Figure AI, Cohere, and Perplexity, among several others.
Beyond the big models: strategic diversification
The bet is not limited to the LLM heavyweights. NVentures diversifies into biotechnology, robotics and AI applied to media and entertainment (fields that will consume enormous amounts of computing) including investments in AI-assisted drug discovery companies and content generation platforms.
There are also regional commitments, such as programs to promote national startup ecosystems.
Hardware-driven innovation or bubble?
The approach offers clear advantages: it accelerates innovation, reduces the time-to-market of computationally demanding projects and consolidates Nvidia’s position as an essential supplier.
But the risks are real: conflicts of interest, distortion of market signals, and the possibility of inflated valuations if much of the spending is directed toward CapEx (capital goods investments) in hardware rather than actual business traction.
Nvidia is buying and architecting its own future: a circle where investment fuels demand for its chips and where dependence on the ecosystem reinforces its dominance. That strategy accelerates AI penetration, but also demands scrutiny.
The line between building a solid platform and fueling a bubble supported by CapEx is thin; Monitoring how this balance evolves will be key to understanding whether this circular empire is sustainable in the long term.
This post is also available in: