Support "new forces," partner with the "national team," NVIDIA is always vigilant towards its "largest customer."

NVIDIA is vigorously promoting a diversification strategy, seeking to reduce its dependence on tech giants such as Microsoft, Amazon, and Google. By establishing "sovereign AI" partnerships with countries like Saudi Arabia and the UAE, as well as supporting "new cloud" platforms like CoreWeave, this chip giant is attempting to create new revenue streams outside of the tech giants.

Global Expansion: NVIDIA's "Sovereign AI" Strategy

This week, NVIDIA reached a multi-billion dollar chip deal with Saudi Arabia's Humain, while the UAE announced coordination with the U.S. government to build one of the world's largest data centers in Abu Dhabi. These Gulf countries plan to construct a massive artificial intelligence infrastructure, becoming significant buyers of NVIDIA chips.

Such "sovereign AI" deals are a key part of NVIDIA's diversified customer strategy. Company executives have stated that multiple national governments have approached NVIDIA, hoping to procure its chips for similar sovereign AI projects. CEO Jensen Huang visited the Gulf region this week to showcase the strategic model the company hopes to replicate globally.

Supporting "New Cloud" Platforms: Competing Head-On with Tech Giants

In addition to expanding government clients, NVIDIA is also actively supporting potential competitors to Amazon AWS, Microsoft Azure, and Google Cloud. This includes incorporating "new cloud" platforms like CoreWeave, Nebius, Crusoe, and Lambda into its expanding "NVIDIA Cloud Partner" network.

These partners gain priority access to NVIDIA's internal resources, such as consulting teams that design and optimize data centers for their specialized equipment. In some cases, NVIDIA has also invested in these new cloud platforms, including CoreWeave and Nebius. In February of this year, NVIDIA announced that CoreWeave became "the first cloud service provider to universally offer the NVIDIA Blackwell platform."

Vipul Ved Prakash, CEO of the new cloud platform Together AI, stated that becoming an NVIDIA cloud partner "provides good access to NVIDIA's organization itself." He added, "If the cloud computing giants ultimately become competitors rather than customers, it will be very important for NVIDIA to have its own cloud ecosystem."

Recently, NVIDIA has also formed alliances with suppliers such as Cisco, Dell, and HP to help sell products to enterprise customers. These enterprises manage their own IT infrastructure rather than outsourcing it to cloud service providers.

Jensen Huang stated in March of this year, "I am now more convinced than a year ago (about business opportunities outside of large cloud service providers)." At this year's GTC conference, when NVIDIA launched the Rubin chip, partners like CoreWeave and Cisco were more prominent than traditional tech giants. Jensen Huang predicts that "every industry" will have its own "AI factory"—facilities dedicated to its powerful chips, representing potential new sales opportunities worth hundreds of billions of dollars.

Challenges and Risks: Counterattacks from Tech Giants

Despite NVIDIA's strong push for a diversification strategy, the company's recent regulatory filings for the last fiscal quarter (ending in January) still warned investors that it remains dependent on "a limited number of customers," who are widely recognized as the tech giants operating the largest cloud services and consumer internet servicesAt the same time, these tech giants are developing their own AI chips and pushing them to customers. Amazon, as the largest cloud service provider, is entering the AI training field currently dominated by NVIDIA. The AI startup Anthropic, in which Amazon has invested, is using AWS Trainium processors to train and run its next-generation models.

For NVIDIA and its investors, the outcome of this diversification battle will determine whether it can maintain strong growth momentum when its dominant position in AI chips is challenged