Goldman raises global data center capacity forecast to 273 GW by 2030

Goldman Sachs has revised its forecast for global data center capacity, projecting a total of 273 gigawatts by 2030. This significant increase, from the previously estimated 217 gigawatts, is attributed to heightened demand for artificial intelligence infrastructure and the anticipation of more project announcements. According to their research note published on November 2, this expansion could necessitate a cumulative capital expenditure of $8.5 trillion over five years, with an estimated cost of $50 billion per gigawatt, encompassing computing infrastructure.

The U.S. capacity forecast has also been raised to 152 gigawatts by 2030, marking a 27-gigawatt increase from the prior projection of 125 gigawatts. This surge in U.S. capacity is projected to account for approximately 60% to 70% of the global additions through 2030. The surge in U.S. forecasts is driven by the growth in cloud, neocloud, and related facilities catering to graphics processing units.

These categories alone added 11.8 gigawatts to the cumulative capacity projections for 2026 through 2030. High-performance computing installations associated with AI workloads are anticipated to lead the construction efforts. It is estimated that projects of at least 75 megawatts could make up 66% of the global additions and 75% of the U.S. additions by 2028.

While Goldman Sachs's projections paint a promising picture for the future of data centers, they also highlight mounting construction risks. These risks include limited power availability, permitting delays, community opposition, and rising financing costs. High-performance computing projects, in particular, rely heavily on debt financing, covering 80% to 90% of the project costs.

Projects that secure electricity supplies and obtain clear regulatory approvals are more likely to proceed successfully, especially as developers face tighter financing conditions and increased scrutiny of new facilities.