Nvidia’s GB300 Chip Fuels Middle East AI Push with US Export Approval
Nvidia’s GB300 chip, part of its Blackwell platform, is set to play a big role in expanding AI capabilities across the Middle East. The US Commerce Department recently approved exports of advanced AI semiconductors to two key players in the region: Saudi Arabia’s HUMAIN and the UAE’s G42. Each firm can now access chips equivalent to 35,000 GB300 processors, a move that underscores the chip’s value for high-powered AI tasks like training massive models and running inferences in data centers.
This approval comes with strict security and reporting rules, reflecting ongoing US concerns about technology flows to regions with ties to China. Still, it’s a clear win for both countries aiming to build their own AI hubs. As Yahoo Finance reports, G42’s CEO Peng Xiao called it a “defining moment,” with plans to invest in local AI infrastructure while giving something back through joint commitments.
What Makes the GB300 Crucial for AI?
The GB300 stands out as Nvidia’s most advanced AI chip, designed to handle the intense compute demands of modern AI workloads, and it has already exceeded sales of previous models like the GB200 as Cantor Fitzgerald highlights. In the broader context of Nvidia’s recent earnings, CEO Jensen Huang noted that demand for Blackwell-generation platforms like the GB300 is “off the charts,” especially for both training and inference in cloud environments. This ties into Nvidia’s fiscal Q3 2025 results, where data center revenue hit $51.2 billion, up 66% from the previous year, crushing expectations and driving record profits Blockonomi, MarketScreener, largely driven by AI needs.
For AI specifically, the GB300’s high computing power enables large-scale deployments that were previously out of reach for many regions. It supports everything from training language models to optimizing real-world simulations, making it ideal for emerging data centers focused on sovereign AI—systems controlled locally rather than relying on US-based clouds.
Middle East Deployments and AI Impact
HUMAIN, Saudi Arabia’s state-backed AI company, is leading the charge with plans to deploy up to 600,000 Nvidia GPUs, including GB300 platforms, over the next three years across Saudi Arabia and the US, expanding its partnership with AWS Ad-hoc-News. According to TS2 Tech, these will power initiatives like HUMAIN Chat, a model targeting over 400 million Arabic speakers using Nvidia’s Nemotron models, and “physical AI” projects with Omniverse for simulating systems.
- Saudi Arabia: Up to 35,000 GB300 chips for massive data centers, including a 500-megawatt facility with Elon Musk’s xAI, potentially using hundreds of thousands of Nvidia GPUs.
- UAE: Similar allocation to G42 for AI data-center projects, helping position Abu Dhabi as a tech hub.
- Regional Scale: Gulf partners aim for up to 1 gigawatt of AI infrastructure by 2030, creating sustained demand for chips like the GB300.
These efforts highlight the GB300’s importance in democratizing AI. Non-US projects like these add to the global demand Nvidia sees, with visibility into $500 billion in revenue from Blackwell and Rubin platforms through 2026. By enabling training and inference at scale, the chip helps countries build AI tools tailored to local needs, from language processing to infrastructure optimization, without full dependence on Western providers.
Overall, this approval isn’t just about hardware—it’s about accelerating AI adoption in fast-growing markets, backed by Nvidia’s dominant position in data center tech.
