The rise of AI is placing immense pressure on network infrastructures, catching many businesses off guard. Despite scaling up compute resources for AI, neocloud providers are facing network bottlenecks. Analyst firm Omdia suggests enterprises scrutinize network capabilities alongside compute power when selecting AI services. Neoclouds, like CoreWeave and Gcore, vary significantly in network strength due to their diverse beginnings.

The study highlights that modern AI performance hinges on seamless data transfer across geographically dispersed systems. With varied origins, some providers’ networks range from basic to sophisticated. As AI adoption surges, neocloud providers must enhance their networking strategies, through partnerships or acquisitions, to stay competitive.

Network resilience, low latency, and secure connectivity have become essential, connecting backbones to edge systems. Camille Mendler from Omdia emphasizes that these are critical for neocloud success as AI workloads shift locations. Global providers like Lumen are urging network upgrades to accommodate AI, likening network function to a nervous system in AI-driven enterprises.

AI’s architecture demands fluid, scalable network infrastructures to support the rapid expansion of digital agents and bots. These entities require networks that can adapt dynamically to handle the tremendous data flow between clouds, datacenters, and edge environments. As AI agents make up over half of today’s internet traffic, networks must evolve to a consumption-based model akin to cloud services to support future technological advancements.