The AI Revolution – a commodities play?

The critical materials that are powering AI

HANEft
(Image credit: HANEft)

A single search query on Chat GPT consumes around 1500% more energy than a simple search google search. The overall energy amounts are marginal on their own. Even taken in aggregate, it is a blip in terms of total global energy demand. However, it is illustrative of the potential big increases in electricity demand that will come from the AI revolution. 

Over the past 20 years, the US has seen its electricity demand stagnate. While its economy has grown, it has been able to avoid boosting its electricity generation thanks to efficiency savings. But this is now changing, and a big reason is the boom in data centre demand – and AI data centre demand in particular

Bloomberg. For illustrative purposes only.

Bloomberg. For illustrative purposes only.

(Image credit: Bloomberg)

For example, the US’ mid-Atlantic region (referred to as PJM) has one of the densest clusters of data centres in the US. Dominion Resources, the utility company servicing the region, is forecasting a large ramp-up in electricity demand over the coming years. This demand increase is principally being driven by data centres which serve advanced computing and AI needs. Similar patterns can be expected across the country. 

Subscribe to MoneyWeek

Subscribe to MoneyWeek today and get your first six magazine issues absolutely FREE

Get 6 issues free
https://cdn.mos.cms.futurecdn.net/flexiimages/mw70aro6gl1676370748.jpg

Sign up to Money Morning

Don't miss the latest investment and personal finances news, market analysis, plus money-saving tips with our free twice-daily newsletter

Don't miss the latest investment and personal finances news, market analysis, plus money-saving tips with our free twice-daily newsletter

Sign up

This trend is also global, with data centre demand expected to double. As the International Energy Agency states, “Data centres are significant drivers of growth in electricity demand in many regions. After globally consuming an estimated 460 terawatt-hours (TWh) in 2022, data centres’ total electricity consumption could reach more than 1 000 TWh in 2026.” The organisation notes that this increase in demand is roughly equivalent to the electricity consumption of Japan. 

So, while many investors often chase the AI theme through exposure to tech stocks, especially through big names such as Microsoft, it is also worth highlighting the materials or commodity angle — a literal picks and shovels approach.

The role of Nuclear Energy

Nuclear energy will potentially play a key role in supplying the electricity for this expected boom in demand, particularly given its zero-carbon credentials. Nuclear energy emits almost no carbon when producing electricity. To meet these growing electricity demands, without derailing the world’s various net-zero targets, nuclear energy is essential.

Sprott Asset Management. For illustrative purposes only

Sprott Asset Management. For illustrative purposes only

(Image credit: Sprott Asset Management)

Technology firms are already noting this. Microsoft is increasingly looking to nuclear energy to meet its data centre electricity needs. In 2023, the tech giant signed an agreement with Constellation Energy to buy nuclear power to cover 35% of the energy needs for one of its Virgina data centres. 

Meanwhile, Amazon Web Services (AWS), the world’s largest operator of data centres, recently acquired a new data centre campus in Pennsylvania. The data centre sits adjacent to, and will be powered by, the Susquehanna Steam Electric Station, the sixth-largest nuclear power plant in the US. 

With more nuclear energy generation, uranium is expected to see greater demand. The uranium market is already tight – primary uranium mine supply is significantly trailing demand, with a cumulative forecasted supply shortfall of approximately 1.5 billion pounds by 2040. This added component will put more pressure on the uranium price, to the potential benefit of the miners.

The importance of copper

But generating electricity is only one part of the story. At the same time, getting the electricity generated by nuclear energy to the end user requires transmission. That requires a lot of copper. A buildout of new data centres will require electricity grid expansions and upgrades. 

As with uranium, the copper market is facing a supply deficit. Copper will be a key metal in the energy transition, with 2.5x more copper wiring in an EV than a conventional car, while solar panels and wind turbines require grid expansions and upgrades. The additional demand for copper from the AI revolution and data centre buildout simply adds to this. 

So, while the growth of AI presents many potential opportunities within technology stocks, that is only part of the story. The expected surge in electricity demand, driven by AI, will add to the investment case for key metals, whether copper for grid expansions and upgrades, or uranium to meet growing lowcarbon energy needs. 

In this anticipated scenario, the miners of these commodities could be poised to benefit. If demand consistently falls short of supply, as forecasts suggest, then miners will need to expand production and extract more of these critical commodities to plug the gap. As the prices of these materials rises, these mining operations should become more profitable, incentivising new miners to enter the field.

To explore HANetf's full range of commodities and mining ETFs, visit our Metals and Mining Hub

Tom Bailey
Contributor

Tom Bailey is Head of ETF Research at HANetf, which entails him staying on top of the various themes and asset classes accessible by ETFs or ETCs on the HANetf platform. Tom was previously ETF Specialist at Interactive Investor, one of the UK's leading direct investment platforms. Prior to this, he was a financial journalist, covering the economy, markets and asset management.