Introduction
In a bold move reshaping the AI infrastructure race, OpenAI and NVIDIA have quietly signed a landmark partnership that will build at least 10 gigawatts of compute power and see NVIDIA invest up to a staggering US $100 billion in OpenAI. This deal is quickly trending because it signals a major bet on massive-scale AI training and deployment—and it highlights how compute is once again centre stage in the generative-AI era. For enterprises, developers and hardware suppliers, the ripple effects could be profound.
Background & Context
The AI boom has been turbo-charged not just by models, but by the compute behind them. OpenAI has grown rapidly into a leading force in generative AI, while NVIDIA has become the hardware powerhouse that fuels many of those advances. For years the two have collaborated; OpenAI previously confirmed that it had “no active plans” to use Google LLC’s TPUs at scale and would continue relying on NVIDIA GPUs.
But now the stakes are far higher—this isn’t just a purchase order. It’s a strategic alignment around infrastructure, investment and long-term compute supply, in a moment when analysts estimate trillions of dollars will flow into AI infrastructure in the coming years.
Key Facts / What Happened
On 22 September 2025, OpenAI and NVIDIA announced a letter of intent to deploy at least 10 gigawatts of NVIDIA systems for OpenAI’s next-generation AI infrastructure.
NVIDIA intends to invest up to US $100 billion in OpenAI progressively, tied to each gigawatt deployed.
The first gigawatt of systems is scheduled for deployment in the second half of 2026, led by NVIDIA’s “Vera Rubin” platform.
OpenAI and NVIDIA will co-optimise their roadmaps: OpenAI’s models and infrastructure software will align with NVIDIA’s hardware and software stack.
Voices & Perspectives
Jensen Huang (Founder & CEO, NVIDIA) said:
“This investment and infrastructure partnership mark the next leap forward — deploying 10 gigawatts to power the next era of intelligence.”
Sam Altman (CEO, OpenAI) added:
“Everything starts with compute… we will utilise what we’re building with NVIDIA to both create new AI breakthroughs and empower people and businesses at scale.”
Analysts view the deal as a game-changer. A recent article noted that NVIDIA’s share price jumped on the news of the $100 billion investment, reflecting how investors see compute-infrastructure as a key battleground.
Implications
For developers and businesses, this deal means one thing: more, bigger, faster AI models—and the infrastructure to support them. Companies building on OpenAI’s models may benefit from improved performance and scale.
For NVIDIA and other hardware vendors, the commitment locks in future demand. It also signals that AI compute is not just a component, but the strategic centre of the race.
On the flip side, the scale of investment raises questions about competition, supply chain risks, energy footprint and geopolitics. If one player controls much of the compute pipeline, issues of access and power concentration emerge.
What’s Next / Future Outlook
Watch for a few key upcoming developments:
The definitive agreement detailing this partnership (terms, timelines, governance) expected in the coming weeks.
The rollout of the first gigawatt of NVIDIA systems in late 2026 will be a milestone to watch.
How this infrastructure gets used: will we see new model announcements from OpenAI, or enterprises gaining access to upgraded services?
How competitors respond: will other large AI firms secure similar compute-scale partnerships, or will hardware suppliers diversify their customers?
Wrap-Up
This alliance between OpenAI and NVIDIA marks a new epoch in the infrastructure game behind AI. It’s not just about models anymore—it’s about the very engines that power them. Keep an eye on how this plays out, because the next front in AI may well be measured in gigawatts.
Top comments (0)