Google Unveils Axion, its First Arm-Based CPU for Data Centres

Google has announced the launch of Google Axion Processors, what it’s calling is a huge advancement in its computing technology, at the Google Cloud Next event held earlier today.

These custom Arm-based CPUs are designed for high performance and energy efficiency in data centres, set to be available to Google Cloud customers later this year.

The Axion CPU is the company’s latest effort in a series of custom silicon developments, following the introduction of five generations of Tensor Processing Units, Video Coding Units (VCUs), and Tensor chips for mobile devices since 2015, such as its Pixel lineup.

The new processors are aimed at addressing the growing demands for general-purpose computing power in fields such as analytics, information retrieval, and machine learning. Google of course will leverage Axion for AI training as it will now compete with Intel, Nvidia and Microsoft in making custom chips for the cloud.

With up to 30% better performance than the fastest existing general-purpose Arm-based cloud instances and up to 50% better performance and 60% better energy efficiency than current x86-based instances, Axion is being touted as a leap forward in computing technology.

Google is already deploying services like BigTable, Spanner, and YouTube Ads on Arm-based servers and plans to extend these deployments with Axion.

The processors are built on Arm’s Neoverse V2 CPU and supported by Titanium, a system of custom silicon microcontrollers designed to enhance performance and efficiency. This technology supports a wide range of general-purpose workloads and introduces Hyperdisk, which is a new block storage service for improved storage I/O processing.

“Google’s announcement of the new Axion CPU marks a significant milestone in delivering custom silicon that is optimized for Google’s infrastructure, and built on our high-performance Arm Neoverse V2 platform. Decades of ecosystem investment, combined with Google’s ongoing innovation and open-source software contributions ensure the best experience for the workloads that matter most to customers running on Arm everywhere,” said Rene Haas, CEO, Arm, in an issued statement.

In addition to its performance features, Axion promises out-of-the-box application compatibility and interoperability, building on Google’s contributions to the Arm ecosystem through the development of Android, Kubernetes, TensorFlow, and the Go language.

With the standard Armv9 architecture, Axion will bring seamless operation of common software on Arm-based servers, bolstering the deployment of Arm workloads on Google Cloud, explains Amin Vahdat, VP/GM, Machine Learning, Systems, and Cloud AI.

Google Cloud customers will soon be able to leverage Axion across various services, including Google Compute Engine and Google Kubernetes Engine, with Arm-compatible software and solutions available on the Google Cloud Marketplace.