Hyperscalers Struggle to Keep Pace with Surging Demand for AI Infrastructure
Goldman Sachs (GS) has issued a report highlighting the surge in demand for artificial intelligence (AI) infrastructure and the challenges that come with it. According to GS, hyperscaler capital expenditure is projected to reach over $760 billion by 2026, at a rate of approximately $2 billion per day.
The AI sector is struggling to build infrastructure fast enough to keep pace with demand, with grid operators facing interconnection requests that were not anticipated in planning models three years ago. Electrical distribution equipment from major manufacturers is also experiencing multiyear backlogs.
As a result, fully islanded data centers are emerging as an alternative for hyperscale AI infrastructure, particularly as grid connection delays increase. GS estimates that one-third of future capacity could be islanded.
The real opportunity lies in bringing AI into the physical world and the broader economy, with the real economy defining the actual scale of opportunity. This includes sectors such as manufacturing, robotics, defense, construction, and energy, which make up approximately 99.5% of the global economy that AI has barely touched.
Dario Amodei, CEO of Anthropic, advocated slowing AI development amid concerns over the risk of losing control of increasingly advanced AI systems.