Nvidia Secures 170,000 GPUs Deal with Firmus, Supports $115B OpenAI Spending
NVDA•Nvidia agreed a strategic partnership with Firmus Technologies to supply 170,000 GPUs to Indonesia from early 2027, potentially driving significant infrastructure product revenue and cloud royalties. Meanwhile, OpenAI’s planned $115 billion infrastructure spending through 2029 and surging AI data center build-out solidify Nvidia’s dominant GPU demand outlook.
1. Firmus Partnership Details
Nvidia and Australian infrastructure provider Firmus Technologies signed a multi-year agreement for the delivery of 170,000 GPUs to a new data center in Batam, Indonesia, scheduled between early 2027 and early 2028. The deal grants Nvidia product sales revenue and cloud service royalties based on GPU utilization by AI Native clients.
2. OpenAI Infrastructure Spending
OpenAI has committed up to $115 billion in computing infrastructure through 2029, with a significant portion earmarked for Nvidia GPUs. This spending underscores Nvidia’s position as the exclusive supplier for AI model training and inference hardware for one of the largest AI platform operators.
3. Data Center Build-Out Outlook
Global AI data center build-out is accelerating as hyperscalers and AI startups alike race to expand capacity. Nvidia’s GPU roadmap and ecosystem enable rapid deployment, with customers including Microsoft, Amazon, Google and emerging regional players leveraging its chips for both cloud and on-premises solutions.
4. Implications for Nvidia Growth
The Firmus partnership and OpenAI’s spending plans could drive sustained GPU demand and revenue growth through 2029, offsetting typical semiconductor cyclicality. Investors will watch order backlogs, margin trends on cloud royalties and Nvidia’s capital expenditure requirements to scale manufacturing capacity.
