Google and Intel Strengthen Their Alliance for AI Infrastructure

Google and Intel recently announced the expansion of their partnership aimed at strengthening Google Cloud's AI infrastructure. This multi-year partnership will allow Google Cloud to continue using Intel processors, particularly the Xeon models, for artificial intelligence, cloud, and inference tasks.

The Xeon processors, especially the latest Xeon 6, are at the heart of this collaboration. Google Cloud has relied on these processors for several decades, and this new phase of partnership aims to deepen this technological relationship.

Additionally, the two companies have decided to broaden their collaboration in the development of custom infrastructure processing units (IPUs). These IPUs are designed to accelerate and manage data center tasks by offloading the CPUs, which is crucial for operational efficiency.

This chip development project began in 2021 and focuses on creating IPUs based on custom ASICs. Although Intel has not disclosed the cost of this agreement, the strategic importance of this collaboration is clear.

This expansion comes amid a strong demand for CPUs within the industry. While GPUs are often used for developing and training AI models, CPUs remain essential for executing these models and for AI infrastructure in general.

Lip-Bu Tan, CEO of Intel, emphasized the importance of this collaboration by stating that artificial intelligence is redefining the construction and evolution of infrastructure. According to him, the scalability of AI requires balanced systems, where CPUs and IPUs play a central role in delivering performance, efficiency, and flexibility.

In this context of increasing CPU shortages, many companies are turning to these chips to meet their technological needs. Meanwhile, Arm Holdings, under the ownership of SoftBank, recently unveiled the Arm AGI CPU, its first internally produced chip, illustrating the global pressure on CPU supply.