NVIDIA Launches Revenue-Sharing Plan

As AI shifts from model development to large-scale inference, demand is moving towards continuously running AI factories that generate tokens at volume.

Meeting this surge requires accelerated, multi-tenant infrastructure that can be deployed quickly, operated at high utilisation and aligned to the economics of token-scale services.

Historically, many emerging AI companies have struggled to access capital-intensive compute, with even long-term agreements often failing to unlock financing for infrastructure.

NVIDIA is responding with a new business model designed to broaden access to compute for a fast-growing ecosystem spanning startups, model builders, enterprises, research organisations and regional AI providers.

Under the approach, AI clouds can acquire NVIDIA infrastructure for AI-native, enterprise and ISV customers via a revenue-sharing and credit-support arrangement. The clouds will sell NVIDIA-powered services, while NVIDIA receives standard product revenue plus a share of cloud revenue on the supported capacity — providing the chipmaker with a recurring, usage-linked earnings stream and accelerating platform adoption among high-growth AI-native customers.

For model builders, inference providers, agent platforms and enterprises scaling AI, the structure promises quicker access to full-stack accelerated computing without waiting for data-centre site selection, power procurement, construction and hardware commissioning.

NVIDIA says capacity is being built around demand, with AI cloud companies standing up DSX AI factories to serve customers and workloads across multiple regions.

Sharon AI is among the first partners and plans to deploy up to 40,000 NVIDIA Grace Blackwell GB300 GPUs. "This strategic collaboration with NVIDIA marks a pivotal moment in Sharon AI's mission to deliver sovereign, large-scale AI compute infrastructure," said James Manning, cofounder and CEO of Sharon AI.

Firmus is developing a DSX AI factory campus in Batam, Indonesia, expected to scale to 360 megawatts and up to 170,000 NVIDIA GPUs. "AI-native companies need access to scalable, energy- and cost-efficient compute infrastructure to compete globally," said Tim Rosenfield, co-CEO of Firmus Technologies. "Firmus AI cloud is building a NVIDIA DSX-aligned AI factory, which will enable our cloud to help more customers access the compute they need to build and scale AI."

AI-native platforms such as Baseten, Fireworks AI and Together AI illustrate where demand is heading: they require immediate access to AI cloud capacity for training, post-training, fine-tuning and high-volume agentic inference. Their customers need dependable large-scale NVIDIA-accelerated compute as usage grows, alongside commercial flexibility as products move from pilot to production.

www.nvidia.com