1. Google has released multiple custom silicon products, including Tensor Processing Units and Video Coding Units
2. The new Axion processor offers up to 30% better performance than current Arm-based instances and up to 50% better performance than x86-based instances
3. Axion processors are built on the Armv9 architecture and offer energy efficiency and application compatibility for a wide range of workloads.
Google has a history of developing custom silicon, with Axion being the latest addition, following previous releases like Tensor Processing Units and Video Coding Units. As customer workloads increasingly require more compute power, investments in accelerators are crucial to keep up with demand and maintain infrastructure efficiency. The Axion processors, utilizing Arm’s CPU cores, offer up to 30% better performance than current Arm-based instances in the cloud and significantly improved energy-efficiency compared to x86-based instances.
Axion processors, underpinned by Google’s Titanium system, provide enhanced performance for various workloads like web servers, databases, AI training, and more. The processors utilize Arm’s Neoverse V2 CPU and are designed to optimize energy-efficiency, aligning with Google’s sustainability goals. Customers can expect compatibility and interoperability with Axion processors, built on the Armv9 architecture, ensuring seamless operation with common software packages and easy deployment on Google Cloud services.
Customers and partners across various industries are excited about the potential of Google’s Axion processors for improving performance, efficiency, and sustainability. From companies like VMware, CrowdStrike, Datadog, Elastic, and WP Engine, organizations are eager to leverage the benefits of the new custom Arm-based CPU. Google Cloud’s commitment to innovation and collaboration within the Arm ecosystem is reflected in the positive responses from partners and the growing interest in deploying workloads on Axion processors.