Jensen Huang, founder and CEO of Nvidia, renewed his expectation that the company’s revenue will grow 70% year over year next year during his participation in the Goldman Sachs Communacopia + Technology conference. Based on estimates that Nvidia will end its current fiscal year with revenue of approximately $400 billion, this growth could raise revenue to around $680 billion.
Huang links this optimism to Nvidia’s position within the artificial intelligence sectors, saying that the models of Anthropic, OpenAI and Google, along with open-weight models, can use the company’s platforms. According to his statements, Nvidia does not merely sell individual chips; it ships large-scale integrated systems that include processors, networking and other data-center components.
From a Graphics Chip to a Complete Infrastructure
Huang cited a system that combines 36 Grace processors with 72 Blackwell graphics processors, saying that sales of this system are currently growing 27% per month. He also described the scale of the systems the company builds, noting that what was previously viewed as a $399 GPU unit has, in one example he presented, become an $8.5 million system comprising two million parts and consuming 250,000 kilowatts, with thousands of units being shipped.
These figures reflect the changing nature of Nvidia’s business: commercial value is no longer tied to the processor alone, but to the computing, networking and power system required to run artificial intelligence workloads. This explains why the company is competing simultaneously with cloud-service providers, laboratories and emerging chip companies.
Why Does This News Matter?
Nvidia faces increasing competition from Amazon, Microsoft and Google, which are developing their own chips, and from laboratories such as Anthropic and OpenAI, in addition to Cerebras and startups such as Etched. Therefore, a 70% growth forecast represents not merely a sales projection, but a bet on the artificial intelligence market’s continued reliance on the infrastructure Nvidia provides despite the expansion of alternatives.
Huang said the company monitors energy, land and data-center facility needs around the world, and that it receives information from cloud providers, original equipment manufacturers and artificial intelligence companies. This reach gives it, according to his interpretation, an early view of expected infrastructure demand.
The Investment Debate and Circular Deals
Huang faced questions about Nvidia’s investments in companies that later purchase its products, arrangements that critics have described as potentially “circular.” He responded that the company invests small amounts compared with the sales that return to it, and said Nvidia verifies before investing that genuine contracts exist that generate revenue from customers. He added that he had seen contracts of this kind worth a combined $100 billion.
However, Huang’s statements remain forecasts from the company’s leadership, not guarantees of future performance. The article itself indicates that artificial intelligence growth currently depends heavily on startups raising substantial funding and spending it on infrastructure, while market maturation could later lead to more efficient use of resources and tokens. The open question, therefore, is whether Nvidia’s current influence will persist when spending intensity declines or alternatives become more capable.