Nvidia’s $1 trillion bet keeps bubble fears at bay
Nvidia has raised its estimate of the potential market for its artificial intelligence chips to $1 trillion by 2027, in a move that helped steady investor concerns about the durability of the AI boom.
Speaking at the company’s developer conference in California, chief executive Jensen Huang said demand for AI computing continues to expand as companies move from building models to deploying them at scale.
“The inference inflection has arrived,” he said, “and demand just keeps on going up”.
The updated forecast, up from a $500bn (£374bn) opportunity outlined earlier this year, comes at a time when markets have begun to question whether spending on AI infrastructure can be sustained at current levels.
Nvidia shares rose modestly by 1.2 per cent following the announcement.
Deployment to meet demand
The company’s latest strategy is focused on so-called inference computing, where AI systems are used in real time to generate responses or carry out tasks.
This differs from the earlier phase of the cycle, when demand was mainly driven by training large models.
Nvidia said its new ‘Vera’ processors would handle the initial stage of processing user inputs, while tech from Groq, licensed in a multibillion-dollar deal, would be used to generate outputs more efficiently.
“We are selling a lot of CPU standalone,” Huang announced. “This is already for sure going to be a multibillion-dollar business for us.”
Recent results suggest underlying demand for AI infrastructure remains robust.
Nvidia has reported record revenue for 11 consecutive quarters, with annual sales reaching $215.9bn.
Data centre chips account for the majority of that growth, reflecting continued investment by large technology companies.
Spending across the sector also remains elevated, with major firms including Amazon, Microsoft and Meta expected to commit significant capital to AI infrastructure this year.
At the same time, usage of AI services continues to expand, causing the likes of OpenAI and Anthropic to report rising revenues as their tools are deployed more widely across businesses.
AI bubble?
Despite that growth, questions remain about how sustainable – or profitable – current levels of investment will prove.
Investors have raised concerns about whether all announced AI infrastructure projects will be completed on time, or deliver the expected returns.
There are also uncertainties around the pace of technological change, where new chip designs and rapid product cycles can affect the long-term value of hardware investments.
Venture capitalist Bill Gurley said this week that a “reset” in the AI market is likely at some stage, as rapid gains often attract additional capital and competition.
“When people get rich quick, a whole bunch of people come in and want to get rich too,” he added.