Nvidia is spearheading the advancement of artificial intelligence within the United Arab Emirates (UAE) by introducing a robust chip and software ecosystem. With plans to release a free AI model, Nvidia aims to bolster its competitive position against leading Chinese technologies.
In a significant move last month, Nvidia invested $7 billion to acquire a stake in the U.S. coding startup Poolside. This investment also includes licensing the company’s AI model-building tools. The anticipated model will be available for free download by the end of the year, primarily targeting users who already operate Nvidia hardware. The UAE is uniquely positioned to leverage this innovation, as its flagship data center will operate using a staggering 400,000 Nvidia chips.
UAE’s Strategic Neutrality
The UAE has strategically positioned itself as a neutral ground between the U.S. and China, sourcing chips from American companies while previously relying on Chinese equipment. To mitigate dependence on outside software, the UAE has developed its own AI model, Falcon. However, in a shift towards closer ties with the U.S., the state-owned AI firm G42 recently severed its connections with Chinese companies to obtain Nvidia chips. If the UAE adopts Nvidia’s new model, it will further solidify this strategic alignment.
Mohammed Soliman, a senior fellow at the Middle East Institute, noted, “The last missing piece was the model layer. Nvidia already controls the chips, and now it is expanding its model factory while releasing open weights.” This means the UAE may not just benefit from Nvidia’s technology but could also become an influential player in AI development.
Build or Lease AI Technology?
The UAE’s two main institutions driving its AI initiatives have differing approaches regarding proprietary model development. The Technology Innovation Institute (TII), responsible for Falcon, advocates for building independent models, prioritizing homegrown solutions. In contrast, G42, which manages the UAE’s AI infrastructure, has been aligning with Nvidia’s initiative to distribute free models.
Hakim Hacid, TII’s chief researcher, emphasized, “The value is not just in the models we produce but in mastering the underlying technology for full control.” This highlights a broader discourse on technological sovereignty, particularly regarding how countries can manage their AI capabilities while remaining open to external innovations.
G42 has tempered its position by actively supporting Nvidia’s push for accessible models, indicating a collaborative approach to AI development. In July, they became the only Gulf firm to join Nvidia’s initiative to ensure countries retain control and transparency over AI models.
Nvidia’s Strategy and Future Aspirations
Nvidia has compelling incentives for encouraging global adoption of its upcoming model, likely named Nemotron 4. Previous U.S. AI models lagged behind China’s offerings, which were downloaded around 2 billion times this year—significantly outpacing Western alternatives. In response, Nvidia is poised to acquire Hugging Face for nearly $13 billion, aiming to position Nemotron as a primary choice for governments building their AI frameworks.
The goal appears to be less about replacing existing models and more about fostering an environment where Nemotron gains traction among users already invested in Nvidia hardware. TII believes it can innovate while learning from global advancements, with Hacid stating, “Sovereignty does not mean isolation. We can build our foundation at home while being receptive to the world’s best technologies.”
As the UAE’s ambitions in AI grow, the choices made now will have lasting implications for its technological landscape and its role as a regional power in AI development.