Nvidia Rolls Out Revenue-Sharing GPU Deals and Predicts $3T–$4T AI Capex by 2030
NVDA•Nvidia forecasts global AI data center capital expenditures will reach $3 trillion–$4 trillion by 2030, underscoring surging GPU demand. It also launched a revenue-sharing program offering token credits to AI startups and is extending financial guarantees to emerging GPU cloud providers in exchange for future revenue.
1. Aggressive AI Capex Forecast
Nvidia projects that global AI data center capital expenditures will total between $3 trillion and $4 trillion by 2030, driven by rapid deployment of H100 and next-generation GPUs. The company sees this surge as a key driver of its data center revenue growth over the coming four years.
2. Startup Revenue-Sharing Program
Nvidia has introduced a new program granting AI startups token credits for GPU access in exchange for a share of their future revenue. The initiative aims to accelerate adoption of Nvidia’s compute platforms among emerging developers by reducing upfront infrastructure costs.
3. GPU Cloud Financing Initiative
In a separate push, Nvidia is offering financial guarantees and infrastructure financing to select GPU cloud providers. In return, the company secures a portion of these providers’ cloud service revenue, strengthening its position in AI infrastructure.
4. Competitive Landscape
Rivals and major cloud players are designing custom AI chips to challenge Nvidia’s dominance, while Anthropic, Google and others explore alternative suppliers. Nvidia’s dual revenue-sharing schemes and capex forecast are intended to reinforce its ecosystem and fend off growing competition.
