Qualcomm Enters AI Data Center Chip Market, Shares Jump on First Customer Announcement
Qualcomm disclosed plans to ship a custom AI data center processor to an unnamed major hyperscaler later in 2026, sending shares higher. The company also authorized a $20 billion share buyback. The move marks Qualcomm's entry into a market long dominated by Nvidia.
Qualcomm shares jumped after the company disclosed plans to ship a custom AI data center processor to an unnamed major hyperscaler later in 2026. The announcement marks Qualcomm's entry into the high-margin inference and training chip market, which has been dominated by Nvidia.
The company also authorized a $20 billion share buyback alongside the chip announcement, signaling confidence in its financial position and future earnings.
Qualcomm did not name the hyperscaler customer, but analysts said the disclosure alone was enough to move the stock. The company has long been known for its mobile chips, which power most of the world's Android smartphones. Moving into data center chips represents a significant expansion of its business.
The AI chip market has been a source of enormous profits for Nvidia, whose GPUs have become the standard for training and running large AI models. But demand has outpaced supply, and hyperscalers are actively looking for alternatives to reduce their dependence on a single supplier.
Qualcomm's entry could help ease GPU shortages and give data center operators more options. The company said its chip is designed for inference workloads, which involve running trained AI models rather than training them from scratch. Inference is a large and growing part of AI computing demand.
The announcement came the same week that Nvidia secured $23 million in incentives from Irving, Texas, to expand its AI supercomputing presence in the United States. Both moves reflect the intense competition for position in the AI infrastructure market.
Analysts said Qualcomm's data center ambitions are credible given its experience designing high-performance chips for mobile devices, which require efficient processing under tight power constraints.
