Nvidia and OpenAI have revealed what they describe as history's most ambitious AI infrastructure venture, a collaboration to establish 10 gigawatts of data centres utilising millions of Nvidia's cutting-edge processors.
The agreement could result in Nvidia investing as much as $100bn (£73.99bn) in Sam Altman's enterprise, whilst OpenAI simultaneously purchases billions of pounds worth of those identical chips, as reported by .
This arrangement has captured attention, with Nvidia - the globe's most valuable corporation boasting a market capitalisation approaching $4.5tn - poised to incrementally deploy capital into OpenAI as each gigawatt of computational capacity becomes operational, commencing with the initial phase in 2026 through its new Vera Rubin platform.
In exchange, OpenAI will bind itself more tightly to Nvidia's ecosystem, depending on its GPUs to develop and operate the forthcoming generation of artificial intelligence systems.
"Everything starts with compute", Altman declared. Nvidia chief Jensen Huang characterised the alliance as "the next leap forward" following ChatGPT, solidifying years of intimate cooperation between the two technology giants.
'Circular dynamics' warnings
Markets responded favourably to the announcement, which propelled Nvidia shares almost four per cent higher on Monday.
Hargreaves Lansdown analysts projected the initiative could yield $500bn in Nvidia revenue if completely executed; 'this move cements Nvidia's position as the undisputed king of AI', remarked Matt Britzman, senior equity analyst. However, others adopted a more cautious stance.
Bernstein's Stacy Rasgon highlighted concerns about "circular dynamics" where Nvidia finances a client who subsequently purchases its own products with that funding.
"The size of the deal will clearly start to raise some questions", he told CNBC.
Investors have already observed this "infinite money loop", as OpenAI balances enormous infrastructure obligations against a business model that remains unprofitable.
Annual revenues are estimated at approximately $13bn, yet the firm has committed 49 per cent of its profits to Microsoft in return for $13bn of support, whilst separately securing a $300bn agreement with Oracle to construct 4.5 gigawatts of data centre capacity through the Stargate initiative.
The scale of the challenge
Beyond these financial engineering obstacles lies an equally formidable physical challenge.
Ten gigawatts of additional data centre capacity matches the power generation of ten nuclear facilities, representing nearly half of all utility-scale electricity generation introduced across the US during the year's opening six months.
Despite relaxed permitting procedures, constructing such capacity could require years, placing additional strain on energy networks already stretched by electrification and renewable integration.
Altman himself acknowledged the "unprecedented infrastructure challenge" would encompass not merely securing sufficient chips, but establishing the power infrastructure necessary to operate them. Meanwhile, the demand side of the equation remains unproven.
OpenAI boasts 700 million weekly users, making ChatGPT the most popular AI product globally, but enterprise adoption has been patchy.
Reports have suggested that many corporate pilots fail to deliver measurable returns, while GPT-5 has drawn a muted response compared with the early hype.
The bigger picture
For Nvidia, the advantages are instant. Each gigawatt deployed translates to tens of billions of pounds in chip sales, strengthening its dominance over the AI supply chain whilst hyperscalers and start-ups experiment with bespoke silicon.
Securing OpenAI as a "preferred strategic compute partner" will help ward off competitors and ensure its GPUs remain the sector benchmark.
For OpenAI, conversely, the deal provides breathing space alongside hardware.
With Microsoft, Oracle, SoftBank and others involved, the firm sits at the heart of an enormous interconnected network of Big Tech and Wall Street capital.
Yet with expenses escalating and limitations looming, the question remains whether this surge can maintain momentum long enough to generate the anticipated profits.
