Artificial intelligence (AI) may get all the glamorous headlines, but every data centre eventually runs into the same decidedly unglamorous problem: the power bill. Servers need enormous amounts of dependable electricity, and demand is climbing quickly enough that an energy source once considered yesterday’s technology suddenly looks rather useful again.
The International Energy Agency expects global data centre electricity consumption to roughly double by 2030 to around 945 terawatt-hours (TWh). Meanwhile, more than 70 gigawatts (GW) of nuclear capacity is already under construction globally, one of the highest levels seen in decades.
And of course, Canada is joining in. But a reactor certainly needs uranium. Yet somebody also has to design the reactor, manufacture components, maintain it, refurbish it, and eventually supply fuel for decades. Investors interested in buying stocks in Canada, therefore, don’t need to place the entire wager on uranium.
Source: Getty Images
CCO
Cameco (TSX:CCO) remains Canada’s nuclear heavyweight. It owns major uranium operations, produces nuclear fuel, and holds 49% of Westinghouse, giving investors exposure to everything from digging uranium out of Saskatchewan to servicing reactors around the world.
Cameco maintained its 2026 attributable uranium-production guidance of 19.5 million to 21.5 million pounds despite temporary disruptions. More importantly, contracts already cover average annual deliveries exceeding 28 million pounds through 2030.
Westinghouse adds another engine. Its technology is used across 57% of the global operating reactor fleet, while its AP1000 pipeline includes as many as 91 potential reactors.
The problem is price. Cameco stock recently traded around 100 times trailing earnings, leaving little room for uranium weakness or disappointing project execution. I’d rather accumulate it gradually, particularly during a stock market correction, than chase every nuclear headline.
ATRL
AtkinsRéalis (TSX:ATRL) may offer the cleaner risk-reward trade. The engineering company owns the exclusive licence to market CANDU reactor technology and earns from reactor design, maintenance, life extensions, and new construction.
Its nuclear business is already enjoying the boom that investors have been waiting for. Second-quarter nuclear revenue jumped 18% to $671.2 million, prompting management to raise its full-year nuclear revenue outlook from approximately $2.5 billion to $2.7 billion.
Nuclear backlog sits at $4.2 billion, while the segment generated an impressive 25.6% adjusted earnings before interest, taxes, depreciation, and amortization (EBITDA) margin on net revenue. Shares have recently traded around 18 times forward earnings, considerably less demanding than Cameco stock.
There’s still risk. Nuclear backlog declined from the prior year, while major projects can be delayed by politics, regulation, and construction schedules. Still, AtkinsRéalis can collect revenue years before a new reactor begins producing power. That’s a useful place to stand when countries start ordering more of them.
NXE
For investors willing to tolerate considerably more excitement, NexGen Energy (TSX:NXE) offers the development-stage option. Its Rook I uranium project in Saskatchewan received its federal construction licence in March, and construction is now underway. At full production, NexGen says the mine could supply more than 20% of current global uranium requirements from one deposit.
That potential comes with a rather large asterisk. NexGen doesn’t yet produce uranium commercially, but plans to raise roughly $1 billion over the coming nine months to help finance Rook I. Financing, construction delays, cost inflation, or future share issuance could all hurt investors.
Its roughly $9.7 billion market capitalization therefore already prices in considerable success. NXE is the one holding I’d keep the smallest, even though successful execution could also create the largest percentage upside.
Bottom line
For most investors, I’d start with AtkinsRéalis. Cameco stock remains the premier long-term nuclear name, though its valuation deserves patience. NexGen offers much more speculative upside as Rook I moves from blueprint to actual mine.
The nuclear revival won’t be finished next quarter. Reactors take years to build and can operate for 60 years or longer. That frustratingly slow timeline is also the opportunity. By the time the world realizes how much uranium, engineering, and reactor capacity it needs, these Canadian companies could already have years of work lined up.
