• Investment in the El Paso, Texas AI data centre has jumped from about $1.5bn to $10bn
  • Site will scale to 1 gigawatt and support Meta’s expanding AI compute demands

What happened: Meta scales Texas AI data centre investment amid AI race

Meta Platforms has increased its planned investment in a West Texas AI data centre to $10bn, up from around $1.5bn, according to CNBC reporting cited by Reuters. The facility is located in El Paso and will become Meta’s 29th data centre globally and its third in Texas. It is designed to support large-scale artificial intelligence workloads, including training and running advanced AI models that require significant computing power. The project reflects Meta’s ongoing shift toward building dedicated infrastructure for AI systems rather than relying solely on external cloud capacity.

The site is expected to reach 1 gigawatt of power capacity and come online around 2028. That scale places it among the largest data centre projects in Meta’s global portfolio. The expansion signals how quickly AI infrastructure demand is rising, as companies race to secure land, electricity and chips needed to support next-generation models.

Also read: Vantage to build $25 billion AI campus in Texas

Also read: Meta outlines next-gen in-house AI chips

Why this is important

The investment highlights how artificial intelligence competition is now tightly linked to physical infrastructure. A 1-gigawatt data centre requires massive electricity supply, advanced cooling systems and long-term grid coordination, making energy access a strategic constraint in AI development.

Across the industry, firms including Amazon, Microsoft and Google are also expanding data centre capacity at record speed. This wave of spending is reshaping energy markets and accelerating pressure on regional grids, particularly in US states like Texas where land and power availability attract hyperscale projects. It also underscores a broader shift: AI leadership increasingly depends on access to compute infrastructure, not just algorithms or software innovation. Companies that secure large-scale capacity early may gain a durable advantage in deploying and monetising AI services globally.