Nvidia founder expects $1 trillion in revenue from AI hardware by the end of 2027

At the opening of GTC 2026, Jensen Huang voiced an ambitious assessment of Nvidia's future sales: by December 31, 2027, the company, according to him, could receive no less than $1 trillion in revenue from artificial intelligence hardware solutions. For the market, this is not just a striking figure, but a marker of how quickly corporations continue to invest in computational infrastructure for training and running AI models.

Forecast from the Stage

The statement was made not at a quarterly earnings report, where investors expect formal guidance, but directly at the opening of GTC 2026. This is precisely what makes Huang's remark particularly revealing: he spoke not in the dry language of a financial document, but in the context of a technology event, where products, architectures, and roadmaps are typically discussed. When the Nvidia CEO delivers such a figure to the main conference stage, it looks like a demonstration of confidence that demand for AI infrastructure will remain very high for at least several more quarters.

In essence, Huang fixed an expectation that in less than two years, Nvidia could collect a trillion dollars specifically from the hardware side of the AI market. This is not about subscriptions, advertising, or cloud services, but about the physical foundation on which modern models operate: accelerators, servers, network components, and related systems. This emphasis is important because it shows where in the chain of AI value creation the largest sums are concentrated today.

Where This Scale Comes From

The reason for such an assessment is simple: major tech companies, cloud providers, and increasingly ordinary corporations continue to build and expand data centers for artificial intelligence tasks. Models are becoming heavier, inference is ceasing to be a peripheral workload, and AI features are turning into a basic part of products. Against this backdrop, Nvidia sells not just a single chip, but an entire stack of computational infrastructure, and this is precisely what allows the company to view the market in terms of hundreds of billions rather than individual successful quarters.

• Accelerators for training and running models
• Server platforms and ready-made racks
• Network solutions for linking large clusters
• Updating existing data centers for AI workloads

An additional factor is the shift in demand from experiments to industrial deployment. If previously many companies tested generative models in pilot mode, now budgets more often go toward long-term purchases. This changes the scale of solutions: instead of a few racks, a business orders entire clusters, and along with them, power supplies, cooling systems, network topology, and capacity reserves for future model releases. For Nvidia, this means not one-off sales, but a sustained wave of infrastructure orders.

Why the Market Listens

Even for Nvidia, a trillion dollars by the end of 2027 is a figure that sounds like a statement about the size of the entire next phase of the AI boom. If this guidance is even close to reality, it means the largest customers are already thinking of infrastructure as a mandatory asset rather than an experimental expense item. This also increases pressure on competitors: all other hardware manufacturers will have to prove that they can offer comparable performance, availability, and ecosystem for corporate clients.

At the same time, it is important to remember that this is an assessment from the company's founder, not a formal forecast from the finance department. The outcome will be influenced by the pace of new capacity deployment, clients' willingness to maintain capital expenditures, supply chain constraints, and how quickly the market will transition from training models to mass inference. But the mere fact that such a figure was publicly named from the GTC 2026 stage already sets a new bar of expectations for the entire AI hardware sector.

What This Means

For the market, the signal is simple: Nvidia believes that peak demand for AI equipment has not yet passed. For business, it means that infrastructure around models remains one of the most expensive and strategically important layers of the entire AI economy.