IA 360
Current Affairs

Nvidia to ship 18,000 GB300 systems to Saudi AI factories

Nvidia will supply HUMAIN with 18,000 GB300 systems in the first phase of a deployment involving hundreds of thousands of GPUs in Saudi Arabia. The deal combines the kingdom’s technological ambitions with Washington’s shift on chip exports.

5 min read Leer en español

Nvidia has struck one of the biggest artificial intelligence infrastructure deals announced to date with Saudi Arabia’s HUMAIN. The company will supply hundreds of thousands of GPUs to Saudi Arabia over the next five years, beginning with 18,000 GB300 Grace Blackwell systems.

The announcement, made on Tuesday during Donald Trump’s visit to Riyadh, matters both for its scale and its geopolitical implications. Saudi Arabia wants to turn its oil revenues into domestic computing capacity. The United States, meanwhile, is easing a regulatory barrier that could have limited precisely these kinds of exports.

An AI factory of up to 500 megawatts

HUMAIN is a new artificial intelligence subsidiary of the Public Investment Fund, Saudi Arabia’s sovereign wealth fund. Its plan is to build so-called AI factories in the kingdom: data centers designed to train, adapt and run AI models at scale, rather than simply host conventional cloud services.

Nvidia’s first delivery will include 18,000 GB300 Grace Blackwell units with InfiniBand networking. GB300 is Nvidia’s latest generation of AI systems, combining its Blackwell Ultra accelerators with Grace processors and high-speed connections to distribute the same workload across thousands of chips.

The full project aims for capacity of up to 500 megawatts. That figure matters because electrical power has become one of the physical constraints in the AI race. Training and serving advanced models requires tens of thousands of GPUs, along with cooling, networking and a stable power supply. Building this infrastructure locally would allow Saudi companies and government agencies to process more data inside the country and reduce their dependence on data centers in the United States or Europe.

The agreement also includes plans for the Saudi Data & AI Authority, the government’s data and AI agency, to deploy up to 5,000 Blackwell GPUs for a sovereign AI factory and smart city applications. Aramco Digital, the technology arm of the state-owned oil company, will also work with Nvidia on computing infrastructure.

Petrodollars enter the race for computing power

Riyadh has spent years trying to diversify an economy heavily dependent on oil through its Vision 2030 program. AI fits that objective in two ways: as an industry that can attract foreign companies and as a technology for modernizing existing sectors, from energy to public administration.

The difference from previous digitalization plans is that Saudi Arabia no longer aims to be merely a customer for US applications. HUMAIN wants to control a more strategic layer: data centers, GPUs and models adapted to Arabic and the needs of local businesses. That does not make the country self-sufficient—its chips, software and much of its expertise are still American—but it does give it greater leverage and operational capacity.

Nvidia is the immediate big winner. Its shares rose 5.6% on Tuesday after the agreements were announced. Jensen Huang accompanied Trump alongside other technology leaders, including OpenAI’s Sam Altman, AMD’s Lisa Su, Elon Musk and Palantir’s Alex Karp.

AMD also announced a collaboration with HUMAIN to deploy up to 500 megawatts of AI capacity over the next five years. The competition will not be limited to selling chips; it will be about becoming the go-to provider for the region’s future sovereign clouds.

Washington opens the door while retaining control

The deal comes on the same day that the US Department of Commerce announced it would rescind the AI diffusion rule approved by the Biden administration in January. The regulation, which was scheduled to take effect on 15 May, divided the world into different levels of access to advanced chips and affected countries such as Saudi Arabia and the United Arab Emirates.

Scrapping the rule makes the political framework for major deals with Gulf partners more accommodating, but it does not eliminate export controls. The United States continues to impose restrictions aimed at preventing advanced chips from reaching China or other destinations deemed sensitive, and its Bureau of Industry and Security has issued new guidance to monitor technology diversions.

That balance will be decisive. For Nvidia, Gulf countries are customers capable of financing enormous data centers at a time when sales of its most advanced products to China are restricted. For Washington, sharing AI capacity with wealthy allies strengthens the US technology ecosystem, but it also requires monitoring who operates these systems, what data they process and how hardware re-exports are prevented.

The Saudi pledge to invest $600 billion in the United States, announced by the White House during the visit, gives the agreement a broader economic framework. The technology deal alone does not amount to that sum, but it shows where interest is concentrating: oil is now financing a race for electricity, data centers and AI chips.

Share this article

Related articles

Apple v. OpenAI, week 2: what's in the case file and what may come this week
Current Affairs

Apple v. OpenAI, week 2: what's in the case file and what may come this week

After Friday's headlines, the docket in case 5:26-cv-07078 spent its first weekend unchanged: the twelve entries are still the ones from day one. We go through them entry by entry — fees, summonses, an eight-lawyer team — and lay out the four procedural milestones that may land this week: the scheduling order, the first proofs of service, the injunction motion Apple has announced, and the July 24 decision on who judges the case.

3 min
Apple sues OpenAI: the partnership that brought ChatGPT to the iPhone ends up in court
Current Affairs

Apple sues OpenAI: the partnership that brought ChatGPT to the iPhone ends up in court

Apple accuses OpenAI, its hardware chief Tang Tan, engineer Chang Liu and io Products of misappropriating iPhone trade secrets to accelerate its first consumer device. The federal lawsuit, filed in San Jose, brings six claims and seeks injunctions and damages; OpenAI replies that it has "no interest in other companies' trade secrets." The case will test where the line sits between hiring talent and taking confidential know-how in the middle of the AI hardware race.

5 min

This website uses cookies to improve the browsing experience. Cookie policy.