Nvidia turmoil and AI data centres: copper demand timing risks for miners
Reviewed by Tom Sullivan

First reported on MINING.com
30 Second Briefing
Nvidia CEO Jensen Huang’s abrupt resignation, amid governance and US national security concerns and a 30% share-price drop to about $148, is testing assumptions that AI-led data-centre growth will drive sustained metals demand. Microsoft is still guiding for about $170 billion of 2026 capex and Alphabet for $195–$205 billion, but with roughly 65–75% of spend going into servers and processors rather than buildings and power systems, any slowdown would mainly hit copper- and aluminium-intensive grid connections, substations and cooling infrastructure. S&P Global projects global copper use rising from 28 Mt in 2025 to 42 Mt in 2040, with AI-related data centres and power adding only ~2 Mt, so miners face more of a timing and financing risk than a wholesale demand collapse.
Technical Brief
- Microsoft spent $45.2 billion in the quarter to 30 September 2026, including finance leases.
- More than $50 billion of Microsoft’s total capex is scheduled just for the December 2026 quarter.
- About three-quarters of Microsoft’s recent technical spend went to shorter-lived processors, memory and storage hardware.
- Alphabet allocated roughly 65% of its technical infrastructure budget to servers and networking, 35% to data-centre facilities.
- S&P Global’s January forecast lifts global copper consumption from 28 Mt (2025) to 42 Mt (2040).
- AI-related data centres and associated power generation contribute only ~2 Mt of that 14 Mt copper increase.
- The IEA projects global data-centre electricity use rising from 485 TWh in 2025 to 950 TWh in 2030, stressing grid and equipment constraints.
Our Take
The projected 2 Mt uplift in copper demand from AI and data centres sits within a much larger 14 Mt increase in total copper consumption between 2025 and 2040 in our database, suggesting AI-linked loads are material but not the dominant driver miners should bank their project pipelines on.
IEA’s forecast that data-centre power use could reach 950 TWh by 2030 aligns with Microsoft’s 20‑year offtake-style deal for the 835 MW Crane plant in Pennsylvania, signalling that long-dated power-securement contracts may become a critical enabling factor for new copper, aluminium and uranium capacity tied to AI infrastructure.
Earlier coverage of Nvidia’s $1 trillion market-cap slide highlighted broad exposure across critical minerals from cobalt and rare earths to tin and hafnium; combining that with this piece’s focus on copper and silver indicates that AI‑hardware cyclicality is now a cross‑commodity risk factor planners may need to stress‑test in project economics.
Prepared by collating external sources, AI-assisted tools, and Geomechanics.io’s proprietary mining database, then reviewed for technical accuracy & edited by our geotechnical team.
Related Articles
Related Industries & Products
Mining
Geotechnical software solutions for mining operations including CMRR analysis, hydrogeological testing, and data management.
Construction
Quality control software for construction companies with material testing, batch tracking, and compliance management.
CMRR-io
Streamline coal mine roof stability assessments with our cloud-based CMRR software featuring automated calculations, multi-scenario analysis, and collaborative workflows.
HYDROGEO-io
Comprehensive hydrogeological testing platform for managing, analysing, and reporting on packer tests, lugeon values, and hydraulic conductivity assessments.
GEODB-io
Centralised geotechnical data management solution for storing, accessing, and analysing all your site investigation and material testing data.


