NVIDIA has launched a new revenue-sharing partnership model with AI cloud providers to accelerate the deployment of large-scale multi-tenant AI factories. This initiative is designed to facilitate faster access to AI resources for startups, model builders, and enterprises, alleviating the need for significant upfront capital expenditures. As part of this collaboration, Sharon AI plans to deploy up to 40,000 Grace Blackwell GPUs, while Firmus is constructing a massive 360MW campus in Indonesia expected to support up to 170,000 GPUs.

Firmus: Firmus develops and operates large-scale data center campuses optimized for AI workloads. Through the NVIDIA partnership, it is advancing construction of a major facility in Indonesia to support extensive GPU-based AI factories. This positions Firmus as a key player in accelerating global AI infrastructure availability.
NVIDIA: NVIDIA develops graphics processing units and AI computing platforms that power data centers and accelerated computing applications. The company is launching a revenue-sharing partnership model with AI cloud providers to enable deployment of large-scale multi-tenant AI factories. This approach directly supports the infrastructure scaling described in the news by reducing barriers for users seeking advanced compute resources.
Sharon AI: Sharon AI operates as an AI-focused cloud provider delivering GPU-accelerated infrastructure services. It is adopting NVIDIA’s new partner model to expand its capacity for large-scale AI deployments. The company’s involvement helps deliver faster access to AI compute for startups, model builders, and enterprises.

Access Benefits: Startups, model builders, and enterprises can obtain large-scale AI resources more quickly through the new partner ecosystem.
Partnership Model: NVIDIA is rolling out a revenue-sharing approach with AI cloud providers to speed up deployment of multi-tenant AI factories.
Infrastructure Acceleration: AI cloud operators are expanding facilities to support broader access to advanced compute without requiring heavy upfront capital from users.