For Your Engineering Success

Spike Talk — A Sustainable “One-Solution-Fits-All” Technology for Electronic Grids

Man,Using,Chatgpt,In,The,Mobile,Phone.,This,App,Is

Power electronics and digitalization, including artificial intelligence (AI) and information and communication technologies (ICT), are crucial for the green transition by optimizing energy efficiency and integrating renewable energy sources into the grid. AI and advanced data analytics enable real-time monitoring and management of energy systems, reducing waste and improving grid stability. Additionally, power electronics facilitate the development of efficient energy storage solutions and smart grids, which are essential for accommodating the fluctuating nature of renewable energy sources.

However, the increased use and the techno-optimism behind Generative AI (Gen-AI) has led to an outburst in carbon footprint due to substantially high energy required for training and running these systems. A comprehensive account of the energy footprint behind using these technologies has been summarized in Figure 1a. Gen-AI has caused a surge in the electricity prices worldwide, with average prices in the U.S. increasing by approximately 15% over the same period, mostly driven by the growing demand for energyintensive technologies, as evident in Figure 1b. Apart from sourcing the servers, a significant amount of power is invested into cooling and data storage, which has gradually caused a large-scale technology-induced shift towards near edge computing. Despite new innovations in the field of energy-efficient computing that reduces the energy consumption of “green” edge AI technologies, we are far away from a system-level integrated framework that can accommodate these technologies sustainably, given that the energy consumption of digital technologies is forecasted to be approximately twice by 2026, as shown in Figure 1c.

https://ieeexplore.ieee.org/document/10839209

Explore More

Latest from the Magazine