NVIDIA Launches Revenue-Sharing Model to Accelerate AI Infrastructure Buildout
Summary
- • NVIDIA introduces a new revenue-sharing partnership model enabling AI clouds to access GPU infrastructure in exchange for sharing cloud service revenue
- • Sharon AI is deploying up to 40,000 NVIDIA Grace Blackwell GB300 GPUs as one of the program's first partners
- • Firmus is building a 360-megawatt AI factory campus in Batam, Indonesia, targeting up to 170,000 NVIDIA GPUs
- • The model directly targets the financing gap that historically blocked AI startups and regional players from large-scale GPU access
Details
Revenue-sharing partnership model
NVIDIA introduces a structure where AI clouds sell NVIDIA-powered services and NVIDIA earns both standard product revenue and a share of cloud service revenue on supported capacity
Usage-linked recurring revenue
The model gives NVIDIA a recurring, usage-linked earnings stream beyond hardware sales, aligning its financial incentives with ongoing AI inference production at partner clouds
Sharon AI: 40,000 GB300 GPUs
Sharon AI is deploying up to 40,000 NVIDIA Grace Blackwell GB300 GPUs as one of the first companies to operate under the new business model
Firmus: 360MW Indonesia campus
Firmus is building a DSX AI factory campus in Batam, Indonesia, expected to scale to 360 megawatts and up to 170,000 NVIDIA GPUs to serve regional AI demand
Democratizing compute access
The model aims to give AI startups and regional players faster access to full-stack compute without delays from site selection, power procurement, construction, and hardware bring-up
Target AI-native ecosystem
Companies like Baseten, Fireworks AI, and Together AI represent the demand side — needing immediate, scalable GPU access for training, fine-tuning, and high-volume agentic inference
DSX AI factory concept
DSX AI factories are multi-tenant, continuously operating accelerated computing facilities optimized for token-scale AI inference economics rather than traditional bursty cloud workloads
Key components of NVIDIA's new AI infrastructure partnership model and its first large-scale deployments
What This Means
NVIDIA's revenue-sharing model marks a strategic evolution from pure hardware vendor to infrastructure platform partner, aligning its financial interests with the ongoing production success of AI clouds rather than just point-in-time hardware sales. By lowering the capital access barrier for emerging AI companies and regional players, NVIDIA is simultaneously expanding its addressable market and deepening structural dependencies on its GPU stack. The scale of initial commitments — up to 210,000 GPUs across just two announced partners — signals this is a serious infrastructure acceleration play with immediate real-world deployment. For the broader AI ecosystem, this could meaningfully speed the global spread of AI compute capacity, particularly in underserved regions like Southeast Asia, while further consolidating NVIDIA's dominance at the hardware layer of the AI stack.