Google unveiled its first custom Arm-based CPUs for data centers at its Google Cloud Next 24 event.
The new Google Axion processors are intended for general-purpose workloads such as web and app servers, containerized microservices, open source databases, and so on.
The company’s investment in custom silicon dates back to 2015, when the tech giant launched its first Tensor Processing Units (TPU). Google has also developed its own Video Coding Unit (VCU) and Tensor chips for mobile devices.
An important milestone
Google’s main rivals in cloud services, Amazon and Microsoft, have their own CPUs based on Arm technology, but Amin Vahdat, Google’s vice president of machine learning, systems and cloud AI boasted: “Axion processors combine Google’s silicon expertise using Arm’s highest-performing CPU cores to deliver instances with up to 30% better performance than the fastest general-purpose Arm-based instances currently available in the cloud.”
Axion CPUs will also have “up to 50% better performance and up to 60% better power efficiency than comparable current-generation x86-based instances,” Vahdat added.
The new processors are built using the Arm Neoverse V2 CPU and the standard Armv9 architecture and instruction set and are backed by Titanium, a system of custom silicon microcontrollers and tiered scale-out offloads, designed to maximize performance for the workload of customers to optimize.
“Google’s announcement of the new Axion CPU marks a significant milestone in delivering custom silicon optimized for Google’s infrastructure and built on our powerful Arm Neoverse V2 platform,” said Rene Haas, CEO of Arm .
“Decades of ecosystem investments, combined with Google’s continued innovation and open-source software contributions, deliver the best experience for the workloads that matter most to customers running on Arm everywhere.”
The contributions to the Arm ecosystem that Haas mentioned include open-source Android, Kubernetes, TensorFlow and the Go language, and should pave the way for Axion’s application compatibility and interoperability. Google says customers can seamlessly deploy Arm workloads to Google Cloud with limited code rewrites, giving them access to an ecosystem of cloud customers and software developers using Arm-native software.
The new Axion processors will be available to Google Cloud customers later this year. Virtual machines based on the CPUs will be available in preview in the coming months.