Qualcomm is once again setting its sights on the data center CPU market, this time teaming up with Saudi Arabia’s (KSA) newly minted AI venture, Humain AI.
After an earlier foray with Centriq processors that stalled in 2019, the chipmaker is leveraging its acquisition of Nuvia’s Arm-based designs to address one of the most pressing challenges in AI: combining raw performance with energy efficiency at scale.
Qualcomm will develop advanced data center CPUs for KSA and other markets
Back in 2017, Qualcomm made headlines with Centriq, its first custom CPU for hyperscale servers. Despite promising benchmarks, the initiative was shelved two years later, leaving Qualcomm’s server ambitions unfulfilled.
In the meantime, the company acquired Nuvia, a startup founded by former Apple silicon engineers, to bolster its Arm-based CPU roadmap. Those designs ultimately found a home in Snapdragon processors for Windows laptops, but speculation persisted that Qualcomm would revive its data center plans once the technology matured.
Qualcomm confirmed on Thursday that it will re-enter the server-CPU market through a memorandum of understanding (MoU) with Humain AI, a state-backed AI cloud infrastructure project funded by Saudi Arabia’s Public Investment Fund (PIF).
Under the agreement, Qualcomm will develop cutting-edge data-center CPUs and support AI accelerators, deploying them in next-generation facilities located first in Saudi Arabia and, potentially, around the globe. In parallel, the chipmaker will establish a Qualcomm Design Center within the Kingdom to localize semiconductor innovation and talent development.
Real-time AI applications, autonomous systems, industrial automation, or smart cities—cannot tolerate round-trip delays to distant cloud servers. By placing AI compute closer to end users (in local data centers or edge nodes), Qualcomm and Humain AI will deliver millisecond-scale responsiveness.
Humain’s roadmap is not limited to domestic facilities. According to the MoU, qualifying data centers could arise in Europe, Asia, and Africa, expanding access to AI infrastructure while fostering a global ecosystem for Arm-based server chips. This comes as the data center market is poised for growth.
Humain has already inked partnerships with Nvidia, AMD, Groq, AWS, and Cisco, signaling broad industry support for its vision.
The partnership will boost Saudi Arabia’s AI ambitions
The Qualcomm-Humain alliance arrives at a pivotal juncture. In early 2025, the US government rescinded the Trump-era AI Diffusion Rule, which had curbed chip exports to many countries. Almost immediately, Nvidia committed to shipping 18,000 GPUs to Saudi Arabia for a planned 500-megawatt data center.
Qualcomm’s move underscores how semiconductor policy and national ambitions are entwining; Saudi Arabia seeks to diversify its economy beyond hydrocarbons by becoming an AI powerhouse, while Qualcomm gains a foothold in a region eager to invest in state-of-the-art computing.
Cristiano Amon, Qualcomm’s President and CEO, remarked that the collaboration will “support the realization” of Saudi Arabia’s global AI aspirations by delivering reliable, cost-effective CPU and AI solutions.
Humain’s CEO, Tareq Amin, described the pact as a “major milestone” toward creating a worldwide hub for AI infrastructure that drives innovation, job creation, and sustainable growth.
For Qualcomm, success hinges on more than just chip performance. It must cultivate software ecosystems, optimising compilers, libraries, and AI frameworks, to unlock the full potential of its custom cores and accelerators. Meanwhile, Humain will need to ensure that its large language models (including Arabic-language variants) are tightly integrated with Qualcomm’s hardware for seamless deployment.
If you're reading this, you’re already ahead. Stay there with our newsletter.
