If chips were the dominant “chips and shovels” trade for the AI revolution so far, energy can be the next. The AI buildout across data centers by hyperscalers requires massive amounts of energy, as evidenced by this report from the International Energy Agency (IEA), which expects electricity consumption to double for data centers by 2030 to a staggering 945 TWh.
And with Meta's (META) sky-high ambitions to build AI superintelligence, it would not be far-fetched to say that the social media giant will duly play its part to augment this energy capacity buildout. To that end, the company recently forged partnerships with some key energy companies in the land for Prometheus, its supercomputing cluster, purpose-built, large-scale system designed to train and run frontier AI models across Meta’s products.
