The rapid growth of AI, particularly agentic AI, is projected to cause a significant increase in data center network traffic, potentially tripling it within three years and increasing by 100x by 2030. Experts at Cisco Live 2026 emphasized that this surge necessitates a "network supercycle," making network infrastructure as critical as GPUs and power for AI workloads. This evolution requires a shift from traditional networking to a three-fabric model: scale-up (within racks), scale-out (across racks), and scale-across (across data centers), with each layer demanding substantially higher bandwidth. AI
IMPACT Data center network infrastructure must evolve to handle massive AI-driven traffic increases, becoming as critical as GPUs and power.
RANK_REASON Expert commentary and projections on the impact of AI on data center networking infrastructure.
Read on Data Center Knowledge →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
