Nvidia explores new AI profit model with startups

Nvidia is exploring a new business model to monetize its AI technology beyond selling chips. The company is looking into offering its AI infrastructure, including its powerful GPUs, as a service directly to AI startups. This would allow startups to access Nvidia's computing power without needing to purchase and maintain expensive hardware themselves.

This move matters because it could significantly broaden Nvidia's revenue streams and deepen its integration into the AI ecosystem. By providing AI infrastructure as a service, Nvidia could capture a share of the operational expenditure of AI model development, rather than just the capital expenditure on hardware. It also positions Nvidia more directly against traditional cloud service providers.

The mechanism involves Nvidia potentially setting up its own cloud-like service or partnering with existing data centers to host its GPUs. Startups would then pay a subscription or usage-based fee to access Nvidia's AI computing resources. This model could lower the barrier to entry for many AI startups, accelerating generative AI adoption and increasing overall AI model development.

This strategy could impact cloud infrastructure spending and affect companies like Amazon (AMZN), Microsoft (MSFT), and Google (GOOGL), which offer similar AI infrastructure services. It also affects other hardware manufacturers who might see Nvidia gain a more dominant position in the AI infrastructure market. Nvidia (NVDA) itself could see increased profitability and market share in the AI sector.