Dedicated stable power support for AI arrival
It is well known that the Artificial Intelligence (AI) requires significant power. In this way, properly designed “Solid Stellar Energy”(SSE) reactor for electrical source can provide significant necessary power to supply the computers (especially the super-computers) in order for the required AI outcomes to be achieved.
Based on the growing use of AI across various industries and the increasing size of models, it’s estimated that AI-related energy consumption in 2024 could range between 20-50 TWh/year, depending on factors like model size, usage patterns, and cloud infrastructure.
By 2030, the global energy consumption of AI could potentially reach anywhere between 100 TWh and 500 TWh per year. The wide range is due to uncertainty around key factors such as hardware improvements, widespread adoption, and the scale of AI model training and deployment. Medium-case estimate: 150-200 TWh per year by 2030, considering moderate growth in AI applications and some efficiency improvements. High-case estimate: 300-500 TWh per year by 2030, if the growth of AI adoption, model complexity, and cloud-based AI services far exceed current trends.