Powering the AI Era: The Critical Role of Copper, Silver, and Energy in Data Centre Infrastructure

We regularly assess structural trends that shape long-term asset allocation for our clients’ portfolios.

The rapid expansion of artificial intelligence (AI) and the associated hyperscale data centres represent one of the most significant demand drivers in the global economy today.

While much attention focuses on computational advancements, the physical foundation of this growth rests on essential natural resources: copper for electrical conductivity and infrastructure, silver for high-performance components, and reliable energy sources to sustain operations.

This report outlines how these resources underpin AI infrastructure, drawing on recent industry analyses and forecasts.

The implications for commodity markets and related investments are substantial, as supply constraints meet accelerating demand.

Copper: The Essential Conductor for AI Infrastructure

Copper remains indispensable due to its superior electrical and thermal conductivity, particularly in high-density environments where space and efficiency are paramount. In data centres optimised for AI workloads, power densities per rack often exceed 20-50 kW, necessitating robust cabling, busbars, connectors, and cooling systems that favour copper over alternatives like aluminium.

Recent projections highlight the scale of this demand. Hyperscale AI facilities can require up to 50,000 tonnes of copper per site, far exceeding traditional data centres.

Global data centre copper consumption is expected to rise significantly, with estimates indicating annual demand from this sector approaching 330,000–420,000 tonnes by 2030 in conservative scenarios, and potentially higher under accelerated AI adoption. AI training clusters alone could account for a substantial portion of incremental use, contributing to broader electrification pressures.

Supply challenges exacerbate this dynamic. Copper faces structural deficits, with new mine development timelines extending 7–10 years.

As data centres proliferate—particularly in North America and Asia—the red metal’s role in power distribution, grid upgrades, and facility build-out positions it as a core enabler of AI scaling.

For sophisticated investors, exposure to copper producers or related equities offers a compelling way to participate in this secular trend.

Silver: Enabling High-Reliability Performance in Advanced Electronics

Silver’s unmatched electrical and thermal conductivity makes it critical for precision components in data centre hardware.

Applications include silver-plated connectors, circuit breakers, relays, semiconductors, and thermal interface materials that ensure reliability under extreme loads and prevent failures in high-uptime environments.

Industrial demand for silver, already dominant, is intensifying with AI infrastructure.

Data centres and related AI systems contribute to growing consumption, alongside solar photovoltaics and electric vehicles.

Forecasts indicate persistent structural deficits through 2030, with U.S. and Chinese data centres potentially requiring hundreds of millions of ounces annually—representing a meaningful share of global mine supply.

Silver’s inclusion on critical minerals lists underscores its strategic importance.

Given stagnant primary production (much as a by-product) and competing uses, silver faces tight availability.

This scarcity supports a constructive long-term outlook for prices and related assets, particularly as AI deployment accelerates.

Energy: The Foundation Sustaining AI Compute

No resource is more fundamental than electricity. Data centres consumed approximately 415 TWh globally in 2024, equivalent to about 1.5% of world electricity use.

The International Energy Agency’s base-case projection sees this doubling to around 945 TWh by 2030, driven primarily by AI-accelerated servers growing at 30% annually.

In the United States, data centres could represent 12% or more of electricity demand by the late 2020s, straining grids and prompting investments in natural gas, renewables, and nuclear.

Hyperscale facilities now demand 100–300 MW or more, with some campuses approaching gigawatt-scale requirements.

This surge not only fuels direct consumption but also necessitates expanded generation and transmission infrastructure.

The interplay is clear: AI’s computational demands drive energy needs, which in turn amplify requirements for copper and silver in power systems.

Diversified exposure to energy infrastructure, utilities, and clean power technologies aligns with this trajectory.

Strategic Implications for Portfolios

These resources are interconnected pillars of the AI economy.

Copper and silver enable efficient, reliable hardware and power delivery, while abundant energy sustains operations.

As AI adoption broadens—from enterprise applications to advanced inference—demand for these inputs will persist, likely outpacing supply growth in the near to medium term.

We view this as a multi-year theme warranting careful consideration in asset allocation.

Opportunities exist across mining equities, commodity funds, energy infrastructure, and related sectors.

We remain focused on high-quality exposures that benefit from these fundamentals focusing on scarcity ahead of abundance.