Intel and Google have expanded their partnership through a new multi-year agreement that focuses on enhancing cloud and AI infrastructure. As part of this collaboration, Google will continue to utilize Intel’s Xeon-based platforms for its upcoming AI and cloud offerings.
The deal involves the deployment of the latest Intel Xeon CPUs alongside custom infrastructure processing units (IPUs), jointly developed by both companies. These IPUs, often referred to as SmartNICs by other suppliers, are designed to alleviate the networking, storage, and security tasks typically handled by CPUs.
Currently, Google Cloud employs Intel processors within its workload-optimized instances, integrating the new Intel Xeon 6 processors into its C4 and N4 instances. This combination aims to cater to a variety of computing needs, from extensive AI training setups to latency-sensitive inference tasks and general-purpose processing.
Lip-Bu Tan, Intel’s CEO, emphasized that effective scaling of AI relies on more than just accelerators; a well-balanced system incorporating CPUs and IPUs is crucial for achieving the performance and flexibility required by modern AI operations.
While Google does offer custom Armv9-based Axion processors aimed at optimizing power efficiency and cost-effectiveness, certain applications require x86 architecture, particularly when high single-threaded performance is paramount.
In a related initiative, Intel has also joined an alliance led by Anthropic, an AI chatbot developer, called Project Glasswing. This coalition comprises over 45 organizations from various sectors, including technology, finance, and cybersecurity, aiming to leverage advanced AI models to identify and resolve crucial software vulnerabilities proactively. The project is built upon Claude Mythos Preview, an emerging AI model by Anthropic, which has already demonstrated its capability by identifying numerous high-severity vulnerabilities across major operating systems and web browsers.
For more information, visit Intel and Google Cloud.
