en.Wedoany.com Reported - WeDoany News, U.S. infrastructure orchestration service provider Rafay Systems announced on May 13 that its platform has officially received NVIDIA AI Cloud-Ready certification. This certification confirms that the Rafay platform meets the software standards set by NVIDIA for AI cloud infrastructure, providing new cloud service providers and sovereign AI operators with the critical capabilities required for a production-grade AI cloud environment. Earning this certification marks Rafay as one of the first independent software vendors to achieve this status, with its platform capable of delivering the API-driven and multi-tenant capabilities needed to serve AI factories at scale.
NVIDIA's AI Cloud-Ready certification runs parallel to its Cloud Partner program, together forming a full-stack standard for AI factory construction. The Cloud Partner program defines reference designs for hardware infrastructure, while the AI Cloud-Ready certification focuses on the software layer, specifying how cloud service providers should deliver and operate AI use cases for customers. Enterprise customers selecting GPU cloud vendors will tend to require their software platforms to meet this standard. Rafay Systems CEO and co-founder Haseeb Budhani stated that this certification, together with the Cloud Partner program, provides a standard for deploying AI factories from hardware to software. By partnering with Rafay, new cloud service providers and sovereign AI cloud providers can offer validated hardware and software from day one of deployment.
The Rafay platform helps operators quickly monetize their infrastructure by providing a full suite of capabilities from bare metal to AI models. The platform supports multiple compute services including Kubernetes, virtual machines, SLURM, and bare metal, and integrates production-grade features such as hard and soft multi-tenancy, quota management, policy governance, lifecycle management, security controls, and self-service. Operators can leverage the platform to offer token-metered model access services, utilizing NVIDIA NIM microservices, NeMo libraries, and AI Blueprints to begin generating revenue as soon as the infrastructure goes live. Budhani noted that when most AI consumers are already paying via tokens, infrastructure providers can no longer simply offer raw GPU rental services.
This certification is the result of years of technical collaboration between Rafay and NVIDIA. The Rafay platform works in concert with the NVIDIA Infra Controller, which handles rack-level configuration for Grace Blackwell systems, while Rafay provides orchestration, governance, and service delivery capabilities on the upper layer. The platform also supports NVIDIA BlueField-3 DPUs and RTX PRO 6000 Blackwell Server Edition, with planned support for BlueField-4 DPUs, ensuring broad hardware compatibility.
In the global market, the Rafay platform has been adopted by multiple operators. Its customers include Yotta in India, Cassava Technologies in Africa, TELUS in Canada, EdgeNet in Latin America, and Firmus Technologies in Australia. Additionally, Rafay has established partnerships with several sovereign cloud operators, including Data Vault in Pakistan, Ooredoo in Qatar, and Moro Hub in the UAE, providing GPU cloud services that comply with local data sovereignty requirements for enterprises and developers in different regions.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
