Saad Ullah
● The world's AI computing power is entering uncharted territory. According to analyst Haider, upcoming projects from OpenAI, xAI, and Anthropic will collectively demand more than 40 gigawatts of electricity before the decade ends — energy levels typically associated with entire industries.
● The numbers are staggering. The Stargate Project will consume 10 GW alone. OpenAI's partnerships with Nvidia and Broadcom will add another 10 GW each, with Broadcom using custom AI chips. OpenAI's collaboration with AMD brings 6 GW more, while Anthropic's million TPUs and xAI's Colossus 2 contribute over 1 GW each. Together, these ventures represent an unprecedented concentration of computing power built to train cutting-edge AI models.
● But this surge brings serious challenges. Experts warn about potential energy shortages, unstable power grids, and massive demand for chips and cooling infrastructure. Without proper planning, this expansion could outrun renewable energy deployment, creating conflicts between data center growth and climate commitments.
● The financial implications are enormous. These projects, worth hundreds of billions collectively, are pushing governments to reconsider energy policies and explore new taxation frameworks. Some are looking at profit taxes or renewable energy credits to balance innovation with environmental responsibility.
Saad Ullah
