The launch of ChatGPT in late 2022 signaled the start of an unparalleled increase in AI-related activity. Although it was just a fundamental generative AI chatbot back then, its monthly visits skyrocketed to 152 million in November of 2022. After that, it reached one billion visits in just three months and was close to four billion by the conclusion of 2024.
However, this marked the beginning of AI’s impact. “The release of generative AI’s second wave in 2025 will generate 100 times more traffic by 2030,” stated Vladimir Galabov, an expert in data centers and co-host of the AIDC Debate podcast. “Due to the growth of agentic AI, it will jump an additional 15 times.” The volume of traffic necessitates a substantial quantity of graphics processing units (GPUs) and immense amounts of energy. An area that is not receiving much attention is the network infrastructure necessary to manage the vast amount of data and speed required for AI tasks.
According to Cisco Chair and CEO Chuck Robbins, speaking at the Cisco Live conference in June 2026, AI is expected to increase networking traffic by three times within three years. This is discussed further in the article titled ‘Networking Strains Under AI’s Scale Demands’.
For those who manage and design data centers, the message is unmistakable.
