Equinix rolls out fabric intelligence for real-time awareness and automation for AI and multicloud workloads.
Data center provider Equinix has launched its Distributed AI infrastructure, which includes a new AI-ready backbone to support high-performance AI deployments spanning multiple data center facilities, an AI Solutions Lab to test new solutions, and Fabric Intelligence to better support next-generation enterprise workloads.
Equinix designed Distributed AI from the ground up to support the scale, speed, and complexity of modern intelligent systems, such as autonomous, agentic AI capable of reasoning, acting, and learning independently.
AI is inherently distributed, drawing on multiple data sources in different locations. To effectively train a model, data must be drawn from multiple locations and processed where it lies, not moved around. This requires a new kind of infrastructure that is globally distributed, deeply interconnected, and fully programmable. Distributed AI links more than 270 data centers in over 60 markets, effectively including almost all Equinix’s facilities, according to the vendor.
“The monolithic kind of enterprise data center is not really a thing that exists anymore,” said Jon Lin, chief business officer at Equinix. “You’ve got workloads that are in SaaS providers. Almost every enterprise is using a multicloud environment. Being able to then to create pipe pipes, to be able to connect all those data flows together is important, because aggregating that at all into one centralized data warehouse is just not really feasible.”
The first part of distributed AI is Fabric Intelligence, a software layer that enhances Equinix Fabric, an on-demand global interconnection platform, with real-time awareness and automation for AI and multicloud workloads. It will be available in Q1 2026.
The second part of the process is in expansion of Equinix’s AI ecosystem, which includes more than 2,000 partners worldwide. The expansion will make next-generation AI inferencing services including the GroqCloud platform discoverable and actionable through the new Fabric Intelligence. It will be available in Q1 2026.
The third part of the offering is the AI Solutions Lab at Equinix Solution Validation Centers. The lab will be available at 20 locations in 10 countries, giving enterprises an environment to collaborate with AI technology partners, test joint solutions, reduce risk and speed up time to market, and collaborate with leading technology partners.
Lin said the current state of affairs has developers doing API calls over the Internet, which is heavily congested with traffic. AI should not have to compete with YouTube videos for bandwidth. “I think that it works by default. It’s able to pass data without a lot of intelligence,” he said. “When you look at the performance characteristics of the Internet, in terms of throughput, as well as potential packet drops, it creates challenges on large scale sets of data flows that might happen.”
