Google and Intel have just announced the expansion of a multi-year strategic partnership to strengthen the capacity of artificial intelligence infrastructure in the context of a strong global demand for AI processing.
According to the new agreement, Google Cloud will continue to use Intel's Xeon processors, including the latest Xeon 6 series, to serve AI, cloud computing and reasoning tasks.
In fact, Google has used Xeon chips for decades, showing the long-standing relationship between the two technology corporations.
Not only stopping at providing hardware, the two sides also expanded cooperation in developing customized infrastructure processing units (IPUs).
This is a specialized chip that helps accelerate and manage tasks in the data center by reducing CPU load, thereby improving overall system performance.
The chip development cooperation program between Google and Intel began in 2021, with a focus on custom ASIC-based IPUs. However, Intel did not disclose the financial terms of the new agreement.
This move comes as the semiconductor industry is facing a shortage of CPUs. While GPUs are often used to train AI models, CPUs play a key role in operating and deploying models in practice, as well as coordinating the entire infrastructure system.
Intel CEO - Mr. Lip-Bu Tan - emphasized that the expansion of AI is not only based on accelerators like GPUs, but also needs a balanced system between many types of chips.
According to him, CPUs and IPUs are core components that help ensure performance, efficiency and flexibility for modern AI workloads.
This trend is also spreading in the industry. Many companies have shifted their focus to CPU production to meet increasing demands.
Recently, Arm Holdings (owned by SoftBank) also introduced the Arm AGI CPU, marking a new step in the global chip race.
The strengthening cooperation between Google and Intel shows that competition in the field of AI infrastructure is entering a new phase, where overall system performance becomes a decisive factor.
