Qualcomm announced plans Monday to launch a new line of custom data center CPUs designed to integrate with Nvidia's artificial intelligence chips, marking the company's return to a sector it exited nearly a decade ago. The move positions the San Diego-based firm to compete more directly with industry heavyweights Intel and AMD, and to capitalize on surging demand for AI infrastructure.
The announcement comes as Nvidia's GPUs remain central to training large-scale AI models, but are typically deployed alongside CPUs. "With the ability to connect our custom processors to Nvidia's rack-scale architecture, we're advancing a shared vision of high-performance energy-efficient computing to the data center," Qualcomm CEO Cristiano Amon said in a presentation at Computex in Taipei.
Amon confirmed the new CPUs would be based on Arm architecture and are part of a long-term roadmap to serve a market he called "a very large addressable market that will that will see a lot of investment for decades to come." While he did not provide specific product launch dates, he said further details on Qualcomm's data center CPU roadmap would be announced "very soon."
The development follows Qualcomm's acquisition of chip startup Nuvia in 2021, which brought in a team of former Apple silicon designers. That acquisition reignited Qualcomm's ambitions in the data center space, which it previously explored in the 2010s with trials involving Meta but ultimately shelved due to cost pressures and legal disputes.
As part of the new strategy, Qualcomm has signed a memorandum of understanding with Saudi AI firm Humain, which is backed by the Public Investment Fund. The agreement calls for collaboration on custom data center CPUs to power AI workloads in the Middle East, further signaling Qualcomm's commitment to diversifying beyond its smartphone-centric legacy.
"The data center will be the fastest-growing segment for the semi market over the next five years," said Mario Morales, group vice president of semiconductors at International Data Corporation. "Companies like Qualcomm haven't benefited from the explosive growth in AI infrastructure because they've lacked a presence. This announcement begins to change that."
While the data center CPU market remains dominated by Intel and AMD, hyperscalers like Amazon and Microsoft have begun deploying custom silicon, narrowing the playing field. Nvidia itself has developed its own Grace CPU using Arm technology, underlining the intensifying competition around AI-capable server hardware.
At Computex, Amon also highlighted Qualcomm's growth in the PC segment, revealing that over 85 PC designs are now in development or on sale using Snapdragon X Series chips. Qualcomm plans to announce a new PC processor at its annual summit in September, reinforcing the company's broader strategy to bring energy-efficient, AI-capable chips to multiple device categories.
