Intel Corporation Expands Partnership with Google Cloud for AI Infrastructure

Intel Corporation is enhancing its collaboration with Google Cloud, with a strong emphasis on AI infrastructure. The focus of this partnership is on developing customised Infrastructure Processing Units (IPUs) and incorporating Intel’s Xeon 6 processors into Google’s data centres.

This partnership is timely, as Intel aims to restore its reputation in a market largely led by NVIDIA. The business is prioritising “heterogeneous” AI systems that rely not only on specialised accelerators but also on CPUs and IPUs, which serve as the “brains” managing extensive AI tasks.

Intel and Google AI CPU Deal

According to Intel’s announcement, this deal will enhance the use of Intel’s Xeon 6 processors within Google Cloud’s data centres, powering various cloud instances such as C4 and N4. These systems are engineered to accommodate AI tasks, including training large AI models, executing AI predictions rapidly, and managing routine computing assignments.

Additionally, Intel and Google will collaborate on developing specific custom chips known as ASIC-based IPUs (Infrastructure Processing Units) to facilitate networking, storage control, and security. As a result, these chips will handle infrastructure functions that were previously the responsibility of CPUs.

CPUs and infrastructure acceleration continue to be fundamental components of AI systems, covering everything from training orchestration to inference and deployment, stated Amin Vahdat, SVP & Chief Technologist, AI Infrastructure at Google.

AI is transforming the way infrastructure is created and scaled, remarked Lip-Bu Tan, CEO of Intel. The scalability of AI requires more than just accelerators; it demands well-balanced systems. CPUs and IPUs are vital in achieving the performance, efficacy, and adaptability that contemporary AI workloads necessitate.

Elon Musk and Intel’s Partnership, and a Larger Manufacturing Commitment

As Intel solidifies its partnerships, the company has also entered into a noteworthy $25 billion agreement with Elon Musk to develop the “Terafab,” a comprehensive chip manufacturing facility in Texas. This facility aims to produce 2-nanometer chips on a large scale, targeting an output of 100,000 wafers each month. This initiative will provide substantial computing support for xAI, Tesla’s autonomous driving technology, and SpaceX.