Google bets on Intel's Xeon 6 CPUs for AI data centers

Google will adopt multiple generations of Intel's central processing units in its artificial intelligence data centers, a significant expansion of their decades-long partnership that positions Intel's new Xeon 6 chip as a key component in future AI workloads. The deal provides a major validation for Intel's turnaround strategy and challenges a market overwhelmingly dominated by Nvidia Corp.

"Their Xeon roadmap gives us confidence that we can continue to meet the growing performance and efficiency demands of our workloads," Amin Vahdat, Google's chief technologist for AI infrastructure, said in a statement.

Under the agreement, for which no financial terms or a specific timeline were disclosed, Google will deploy Intel's latest Xeon 6 processors for both AI training and inference tasks. The move comes as the role of the CPU in AI systems is being re-evaluated; Nvidia's own head of AI infrastructure, Dion Harris, noted in March that CPUs are "becoming the bottleneck" as more complex AI agent workloads emerge.

This partnership is a crucial victory for Intel as it seeks to regain its footing in the high-margin data center market. While Google has developed its own custom AI accelerators (TPUs) for over a decade and recently launched its own Arm-based CPU, Axion, this commitment to the Xeon line underscores a strategy of hardware diversification to mitigate supply chain risks and optimize performance across different AI tasks.

Intel's Turnaround Gains a Key Ally

The Google deal lands amid a series of positive developments for Intel that have helped its shares nearly triple over the past year. The chipmaker secured a 10% stake from the U.S. government and a separate $5 billion investment from rival Nvidia, bolstering confidence in its ambitious plan to build out advanced domestic manufacturing capabilities. The latest Xeon 6 processors are manufactured using Intel's most advanced 18A process technology at its Arizona fabrication plant.

"Scaling AI requires more than accelerators — it requires balanced systems," Intel CEO Lip-Bu Tan said, highlighting the CPU's integral role. This sentiment is apparently shared by other tech leaders, with Tan recently revealing that Elon Musk has enlisted Intel to produce custom chips for his ventures, including SpaceX, xAI, and Tesla.

Deepening a Decades-Long Collaboration

Beyond the new CPU agreement, Google and Intel also reiterated their ongoing collaboration on Infrastructure Processing Units (IPUs), which began in 2022. These specialized chips are designed to offload and accelerate networking, storage, and security functions from the main CPU. According to Google, this allows for more efficient use of compute resources by freeing up the CPU to focus on primary workloads.

For investors, Google's decision to integrate Xeon 6 provides a powerful endorsement of Intel's product roadmap and its 18A manufacturing process. While Intel still faces intense competition from Nvidia's GPU dominance and the growing ecosystem of Arm-based processors, this high-profile design win from a hyperscale cloud leader is a significant step in its effort to become a major force in the AI hardware landscape.

This article is for informational purposes only and does not constitute investment advice.