The growing demand for AI applications, particularly those that require real-time data processing and high-bandwidth communication (such as smart factories, ports, and retail stores), will place a significant strain on existing network infrastructure. Enhancing the energy efficiency of AI is not only vital for environmental sustainability but also for the economic viability of AI applications, particularly in energy-constrained environments.
To address these challenges, we may see increased demand for edge computing, where data processing and analysis occur closer to the source of data generation.