Equinix announced a new solution to accelerate enterprise AI inference through expanded collaboration with NVIDIA and Together AI. Equinix Inference Exchange combines NVIDIA Enterprise Reference Architectures with Together AI’s platform and will be delivered through a global infrastructure.
Equinix announced that it will launch Equinix Inference Exchange, significantly expanding its longstanding collaboration with NVIDIA to accelerate distributed AI inference for enterprises. It will also launch a new collaboration with Together AI. The announcements were made at Equinix Horizon, held in the United States.
The solution integrates NVIDIA Enterprise Reference Architectures with Together AI’s inference platform, providing low-latency, secure connections to data, users and ecosystems through Equinix IBX data centers and Equinix Fabric. Connectivity to cloud and AI providers is also supported.
The inference environment is distributed, enabling metro-edge inference and facilitating the migration of open models. Together AI’s open-model platform can be accessed through existing interconnection infrastructure, with sovereign AI environments also envisaged to meet data residency requirements.
Equinix Inference Exchange is scheduled to become available in the first quarter of 2027, using connectivity on Equinix Fabric and Equinix’s global data center network. Equinix says it will support the operation of distributed inference, backed by more than 280 data centers in 77 metropolitan areas worldwide, 230 cloud on-ramps and an AI ecosystem connected by more than 10,500 enterprises.