Trending
Written by: Renato Monzon, MBA | AGF Investments
Most consumers experience AI through chatbots on their devices. What is easy to forget is that every query they ask their preferred model is answered inside a physical building by complex machines that are powered by electricity.
The result is the demand for power in the developed world rising exponentially after stagnating for two decades. Data centres used about 415 terawatt-hours of electricity in 2024 – roughly 1.5% of the world’s total – and that figure is expected to double by 2030. Current grids are outdated, with new ones taking years to approve and build.
Annual global spending on data centers, meanwhile, is topping a trillion dollars, with cumulative outlays expected to reach $7 trillion by 2030. The capital investment and energy consumption required to support AI is staggering.
AI being built on this physical backbone means industrial companies are benefitting from this rapid infrastructure buildout on multiple levels:
- Buildings: the data centres that house the machines
- Power: electricity generation and transmission through the grid
- Cooling: the systems that keep AI chips from overheating
The money won’t dry up once all this new equipment is installed, either. This new infrastructure will require decades of servicing, maintenance, and upgrades, creating a flow of recurring revenue behind the initial infrastructure investment.
The tricky part for investors is that share prices already reflect much of this optimism; positioning is crowded and valuations are stretched. This pricing in of expected success means that strong results often aren’t enough to drive stocks higher.
The winning stocks are likely to be those that consistently beat high expectations by efficiently turning their growing order books into real cash and profits. Simply meeting guidance may not be good enough.
In summary, while the technological side of AI is likely to keep dominating the conversation, attractive opportunities also exist in the industrial side of the supply chain as the infrastructure needed to power this AI buildout and keep everything running for years to come will be a multi-year opportunity for companies and investors.