Equinix launches AI platform to simplify control of distributed AI resources

An age-old problem for enterprise IT managers has always been data sprawl. However, in the era of AI, where data is needed from every potential source available, scale in data sprawl become unmanageable. Existing architectures weren’t designed for distributing processing, which is part and parcel of AI training. Data center provider Equinix says it has the solution: The Distributed AI Hub powered by Equinix Fabric Intelligence. The platform offers a unified framework offering connectivity across locations to bring together data center, edge, or clouds and AI models or resources. “Every enterprise has come to the realization that AI is not centralized,” said Arun Dev, vice president and global head of Digital Interconnection at Equinix. “What’s becoming more important is inference at the edge, agentic, and this is where the conversations with our customers led us to really look at this distributed AI framework, where they were struggling with a few things.” For starters, customers tell Equinix that they’ve got data in multiple public clouds, Equinix colocation data centers, on prem locations and neo-clouds and other data platforms. In this scenario, there’s concerns around how customers get visibility across all of these different sources of data that are in different locations, according to Dev. There are about 3,000 cloud and IT service providers available through Equinix, Dev said. That includes major hyperscale cloud providers, tier two and smaller clouds, all of them have a presence. The framework builds connections between all of the service providers as well as colocation and on-premises systems. The AI Hub includes a new AI-ready backbone to support distributed AI deployments, a global AI Solutions Lab to test new solutions, and Fabric Intelligence to better support next-generation workloads for enterprises. Fabric Intelligence is a software layer that enhances Equinix Fabric, the company’s on-demand global interconnection service, with real-time awareness and automation for AI and multicloud workloads. It is integrated with AI orchestration tools to automate connectivity decisions, taps into live telemetry for deep observability, and dynamically adjusts routing and segmentation to optimize performance and simplify network operations. To support customers with their AI migration, Equinix is launching a global AI Solutions Lab across 20 locations in 10 countries to give enterprises an environment to collaborate with leading AI partners. Dev said the lab has several customers that are “dipping their toes” when it comes to AI and Equinix helps them validate AI architectures and AI technologies. “The solution validation center is where [pilot programs] become real, and it gives customers the confidence to be able to deploy play in this environment before they move off and actually implement it,” he said. The AI Hub basically provides the orchestration layer for Equinix’s Distributed AI infrastructure announced last fall that provides the physical network for the platform. Dev said the hub is a reference framework, so if customers want to run virtual instances at Equinix, or if customers want to deploy this in their own colocation environment, they’re free to do either. Or customers can take the framework and modify it to their own specific needs. Equinix announced too that the AI Hub would integrate security support from Palo Alto Networks and its Prisma AIRS real-time AI security and centralized policy enforcement package. The idea is to bring real-time AI security and centralized policy enforcement across any location. The Prisma AIRS package will be available on Equinix Network Edge, letting organizations centrally manage AI-driven security services at the digital edge, closer to users, clouds and critical workloads, according to Equinix.