Nvidia's stock price declined on Monday as investors absorbed recent advisories from artificial intelligence leaders regarding the pace of AI development. Nvidia chief executive Jensen Huang reiterated his projection that the artificial intelligence infrastructure sector could reach a value of between $3 trillion and $4 trillion by 2030.

He maintained his position without adjustment, citing figures from another strong quarter alongside the view that demand continues to exceed available supply. The statement was conveyed by Huang at a recent public appearance, and his remarks were covered in full by a broadcaster.

Nvidia's most recent earnings figures were cited as evidence. Quarterly earnings came in at $96.2 billion, surpassing analyst forecasts. Revenue from data-centre operations reached $89 billion, and the company reported adjusted earnings of $2.22 per share.

Supply constraints on memory chips, rather than weakening demand, continue to act as the primary limiting factor, while Huang projected that the semiconductor sector would expand continuously over time. Forecasters had projected 45% expansion in revenue for the 2028 fiscal year. Nvidia has issued updated guidance projecting a 70% figure.

Huang has projected that the proportion could surpass 100% under conditions of unrestricted chip supply, adding at the Goldman Sachs Communacopia + Technology conference — as reported by a news outlet — that Nvidia's processors are used across all models. Each laboratory. Leading artificial intelligence developers, including OpenAI, Anthropic, and Google, are all involved. He contends that this perspective positions him to forecast the next wave with a degree of certainty.

A number of commentators that weekend called for a reduced pace of development in cutting-edge artificial intelligence. Anthropic's chief executive, Dario Amodei, has advocated for a slower pace of development in the field. Both Elon Musk and Sam Altman expressed their backing for the matter. Issues raised spanned employment displacement and the possibility of severe long-term harms.

Semiconductor equities declined as broader trading activity shifted. The PHLX semiconductor index fell by close to 6 percent. Nvidia stock fell by over 3 per cent.

Huang dismissed certain forecasts as unfounded, and at the same Goldman conference, he argued that the volume of cybersecurity discussion is commercially motivated — contending that industry players amplify threat narratives ahead of product launches to generate demand. His valuation range of $3 trillion to $4 trillion, first outlined over a year ago, has remained unchanged despite fluctuating quarterly results and broader economic uncertainty. Huang characterises the expansion as unavoidable.

Firms are working to develop more capable AI systems while the computational demands of running those models continue to expand. Autonomous systems and robotics may increase computational demand by a factor of five to ten over the coming years.

Huang has described his approach as shaped in part by physical infrastructure considerations, including pre-built data-centre structures. The company's business extends well beyond semiconductor hardware. The organisation provides funding for developments and underwrites rental agreements. He told the conference that the platform delivers supply chain visibility that rival products do not offer, while asserting that each unit of investment generates returns one hundred times over — a ratio he noted might appear to favour his own interests.

On Monday, Huang received a speakerphone call from President Trump during the All-In Summit in Los Angeles. Trump characterised AI risk advisories as fraudulent, while Huang indicated the president had correctly identified what he viewed as misleading narratives on the subject. The interaction was covered by a broadcast outlet.

Revenue at Nvidia doubled on a year-over-year basis during the company's most recent fiscal second quarter, climbing 106%. Revenue from data centres rose 117 per cent. The edge computing segment, encompassing gaming and physical AI platforms, also surpassed analyst expectations.

Projected revenue for the quarter ahead stands at approximately $108 billion, with a margin of two percentage points in either direction. The company set a gross margin objective of 74 per cent.

Huang monitors power projects at the gigawatt scale across multiple operators, and the agreements remain in effect. Purchasing activity appears to have moderated rather than expanded speculatively, with one individual noting a preference for guaranteed outcomes.

Huang has outlined a vision that extends well beyond current data-centre investment, encompassing AI agents numbering in the billions alongside physical robotics. Both units draw on computational power continuously. Automated systems are increasingly being developed to construct and refine themselves, displacing conventional approaches to software engineering.

The chip-making industry is expected to scale up substantially to handle that level of demand. Huang has characterised the transition as machines built to serve other machines. That outlook is reflected in announced partnerships, memory agreements, proprietary chip development plans, and supercomputer rollouts.

Huang has stated that certain benchmarks for artificial general intelligence have already been reached in practical applications. His perspective on generative AI is that it currently offers practical utility. Agentic systems are seeing growing uptake across a range of sectors. The uptake of that technology underpins the expenditure he anticipates through the remainder of the decade.

Limitations on available memory remain a relevant constraint, described by Huang as representing confirmed commitments and active initiatives. Available supplies of high-bandwidth memory fall short of what is needed to fulfil all requested configurations. Should those limitations be resolved, production growth could surpass even the 70% threshold he has put forward as a benchmark.

Upcoming product launches, expanded robotics trials, and AI-driven cybersecurity platforms — a domain Huang identified as an area of growing focus — will each serve as tests of whether that $4 trillion valuation holds up.