Equinix, NVIDIA, Together AI는 상호 연결된 콜로케이션 시설과 GPU 클라우드 네트워크 전반에 분산형 엔터프라이즈 추론 워크로드를 배포하기 위한 공동 이니셔티브를 발표했습니다.
Equinix is expanding its collaboration with NVIDIA and partnering with Together AI to launch the Equinix Inference Exchange, a distributed AI inference program designed to help enterprises deploy workloads closer to their data, users, and applications.
The offering integrates NVIDIA’s Enterprise Reference Architectures with Together AI’s inference platform—which supports more than 200 open-source models—and Equinix’s global data centre and interconnection infrastructure. According to the companies, this approach provides enterprises with greater flexibility over where AI inference runs while addressing performance, cost, and governance requirements.
“AI is transforming enterprise technology at extraordinary speed, and the infrastructure decisions enterprises make today will define their competitive position for years to come,” said Adaire Fox-Martin, CEO and president of Equinix. Fox-Martin noted that Equinix’s partnership with NVIDIA provides the accelerated computing foundation for the initiative, while Together AI’s open-model approach offers enterprises greater choice in how they scale AI workloads.
NVIDIA vice president of global AI clouds and infrastructure ecosystem Raj Mirpuri added that the exchange enables organizations to bring AI inference closer to their data, applications, and customers. The program will leverage Equinix Fabric to connect inference providers with clouds, networks, and AI providers across major metropolitan areas, potentially reducing both latency and deployment complexity.
The Inference Exchange will combine Equinix’s infrastructure and operations with NVIDIA’s AI infrastructure and Together AI’s inference platform. Together AI will support both multitenant environments and dedicated single-tenant deployments for organizations requiring exclusive capacity.
Together AI CEO Vipul Ved Prakash stated that the collaboration demonstrates that model choice and performance do not have to be trade-offs for enterprises. The Equinix Inference Exchange is scheduled to become available in the first quarter of 2027.