Equinix launches AI inference & connectivity services

Equinix has introduced two services for enterprise AI infrastructure, focused on AI inference deployment and connectivity across distributed environments.

The first, Equinix Inference Exchange, was developed with NVIDIA and Together AI. The second, Equinix Fabric One, is a managed connectivity service built on open specifications developed by AWS and Google Cloud.

The announcements reflect a shift in enterprise AI spending away from simple access to computing resources and toward practical questions such as where workloads run, how they connect to data and applications, and how companies manage those systems across multiple locations and providers.

Inference Exchange is intended to help companies place AI inference closer to users, applications and data sources. It combines NVIDIA Enterprise Reference Architectures, Together AI's inference platform, and Equinix's global data centre footprint and interconnection network.

According to Equinix, Together AI's platform supports more than 200 open-source models. The service is aimed at companies that want to run inference across different geographies, use open-source models, or meet data residency requirements in regulated industries.

The product will be delivered through Equinix data centres and connected through Equinix Fabric to clouds, networks and AI providers. Equinix said this would give customers a shorter path from AI testing to production deployment.

"AI is transforming enterprise technology at extraordinary speed, and the infrastructure decisions enterprises make today will define their competitive position for years to come. Equinix is uniquely positioned to deliver what this moment demands based on our nearly three decades building the trusted exchange where the world's enterprises run, connect and orchestrate their most critical workloads," said Adaire Fox-Martin, Chief Executive Officer and President of Equinix.

"Our longtime relationship with NVIDIA delivers the accelerated computing foundation at the heart of modern AI, while Together AI's commitment to open ecosystems gives enterprises the flexibility to scale on their terms. Equinix Inference Exchange will enable architectures that are neutral by design, open by default and engineered for exceptional performance," Fox-Martin said.

NVIDIA described the move as part of a broader effort to distribute AI infrastructure rather than centralise it in a small number of locations.

"Equinix Inference Exchange turns the world's leading digital interconnection platform into a global fabric for AI inference," said Raj Mirpuri, Vice President of Global AI Clouds and Infrastructure Ecosystem at NVIDIA. "As accelerated compute becomes a strategic asset class, combining NVIDIA's infrastructure & technology with Together AI's open-model inference platform and Equinix's global reach gives enterprises a powerful, distributed foundation to bring intelligence closer to their data, applications and customers, accelerating the next generation of intelligent services."

Together AI said the arrangement would give customers access to open-source model deployment without relying solely on proprietary providers.

"Together AI was built on the conviction that open, accessible AI is what will define the industry moving forward, because enterprises shouldn't have to choose between model performance and operational flexibility," said Vipul Ved Prakash, Co-Founder and Chief Executive Officer of Together AI. "What we are building with Equinix and NVIDIA proves that model choice and performance are not trade-offs. They are the foundation of enterprise AI done right."

Connectivity layer

The second announcement centres on networking rather than model execution. Fabric One is designed to let enterprises define what should be connected across AI, cloud and enterprise systems, while Equinix manages routing, cloud connectivity, encryption, resiliency and failover behind the scenes.

The service uses the OpenAPI 3.0 Interconnect specification, developed in collaboration by AWS and Google Cloud, to support interoperability across distributed environments. According to Equinix, customers will be able to make requests through a portal, APIs, automation workflows, agent-based requests or natural-language prompts.

That addresses a common problem for businesses adopting AI and multicloud systems. Each new connection between clouds, model providers, data sources and sites often becomes a separate networking project, requiring specialist staff and ongoing management.

"For decades, enterprise networks have been built one connection at a time for each partner and provider they depend on. That approach doesn't scale in a world of distributed AI that demands dynamic, flexible and real-time connectivity," said Chris Audie, Chief Product Officer of Equinix. "Equinix Fabric One will reduce the complexity that's slowing enterprises down by automatically managing connectivity across distributed architectures, based on the customer's intent."

Industry analysts cited by Equinix said enterprises are increasingly seeking ways to automate provisioning and management as cloud and AI environments spread across more providers and locations.

Equinix is positioning both products around its role as an interconnection provider. It says it operates more than 280 data centres across 77 metros, with 230 cloud on-ramps and more than 10,500 interconnected businesses on its exchange. The company also said eight of the top 10 AI model providers and nine of the top 10 AI clouds are deployed with Equinix.

AWS and Google Cloud both backed the use of open interconnect standards in the new service.

"Customers shouldn't have to figure out how to connect across multiple clouds and AI environments on their own," said Robert Kennedy, Vice President of AWS Network Services. "They should just define what they need and have it work. Fabric One builds on the open interconnect specification that AWS and Google Cloud helped define and delivers exactly that: a managed service that handles the complexity so customers can focus on their AI workloads, not how everything connects."

"Cross-cloud interoperability is the backbone of the AI era," said Rob Enns, Vice President of Engineering, Cloud Networking, Google Cloud. "We helped establish the open interconnect specification so enterprises no longer have to manage the complexity of multi-cloud networking. By combining Google Cloud's Cross-Cloud Network capabilities with Equinix Fabric One, we are giving customers an automated, intent-driven network that lets them quickly deploy distributed AI workloads."

Equinix said Inference Exchange will be available in the first quarter of 2027. Fabric One is expected to enter beta later this year, with general availability in 2027 and an initial launch in North America.