Nvidia CEO Jensen Huang recently said he expects Nvidia to sell roughly twice as many chips next year. That is a volume forecast, not a promise that revenue will double, but it shows how strong the demand for AI compute remains. Nvidia has also projected around 70% revenue growth for its fiscal year ending January 2028.
The interesting part is what happens beyond GPUs.
The IEA says electricity consumption from AIfocused data centres jumped 50% in 2025, while total data-centre electricity use is projected to almost double from 485 TWh in 2025 to about 950 TWh by 2030.
That creates several potential AI investment layers: power generation, electrical equipment, grid infrastructure, cooling, networking, optical connectivity and data-centre construction. Reuters recently highlighted strong demand for power and cooling equipment, while the global AI buildout is also driving demand for high-speed optical networks.
But there is a risk investors should not ignore. Big Tech’s AI infrastructure commitments are enormous, and Reuters has estimated roughly $1.09 trillion in future data-centre lease payments across major technology companies. Texas has even paused new state data-centre permits while it audits grid impact, showing that power and regulation can become real bottlenecks.
My view: bullish on the long-term AI buildout, but not every AI stock deserves the same valuation. The next phase may reward companies selling the “picks and shovels” — chips, electricity, cooling, networking and infrastructure — rather than simply chasing whichever AI stock is moving fastest.
The AI boom is real. The harder question is who captures the economics.
Are you bullish or bearish on AI stocks from here? Share your AI-related trade or holdings with the trade widget.
