Equinix turns the network into the control plane for enterprise AI inference
At its first Horizon customer and partner event this week, Equinix Inc. argued that the architecture of enterprise artificial intelligence is being reshaped by a simple yet hard-to-answer question: Where should inference run? For the past several years, much of the AI infrastructure conversation has centered on the supply and cost of accelerated computing. The […] The post Equinix turns the…
Equinix announced two new offerings at its Horizon event that aim to transform enterprises' approach to AI inference. The company's Fabric One service is a managed, intent-driven connectivity platform designed to simplify the complex task of connecting distributed AI environments. Rather than requiring customers to manually configure various network services, Fabric One will compose the necessary connectivity based on specified business or application outcomes.
This shift represents a significant move away from traditional, project-oriented networking models and into a more adaptive control plane for AI infrastructure. The Inference Exchange, built in partnership with Nvidia and Together AI, is a distributed inference offering that brings production-grade inference capacity closer to enterprises' data sources, applications, and users.
This service, which will support both multitenant and dedicated single-tenant environments, addresses the unique challenges of inference computing, such as latency, data movement costs, and varying data sovereignty requirements. CEO Adaire Fox-Martin emphasized that enterprise IT must now contend with fragmented compute resources, the rise of machine-to-machine traffic, unpredictable AI returns, and increasingly stringent data sovereignty concerns.
Both Fabric One and Inference Exchange represent Equinix's effort to extend its interconnection capabilities into the AI era, providing enterprises with a more integrated, flexible, and optimized approach to deploying and operating AI inference workloads.
Written by urgent.news from SiliconANGLE's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
