Jump to Content
Compute

Introducing Google Axion Processors, our new Arm-based CPUs

April 9, 2024
https://storage.googleapis.com/gweb-cloudblog-publish/images/Next24_Blog_Images_6-05.max-2500x2500.jpg
Amin Vahdat

VP/GM, Machine Learning, Systems, and Cloud AI, Google Cloud

Try Gemini 1.5 models

Google's most advanced multimodal models in Vertex AI

Try it

Editor’s note: The first Google Axion Processor is now GA: Check out C4A, our custom Arm-based VMs


At Google, we constantly push the boundaries of computing, exploring what is possible for grand challenges ranging from information retrieval, global video distribution, and of course generative AI. Doing so requires rethinking systems design in deep collaboration with service developers. This rethinking has resulted in our significant investment in custom silicon. Today, we are thrilled to announce the latest incarnation of this work: Google Axion Processors, our first custom Arm®-based CPUs designed for the data center. Axion delivers industry-leading performance and energy efficiency and will be available to Google Cloud customers later this year.

https://storage.googleapis.com/gweb-cloudblog-publish/images/image1_Ab4940U.max-2000x2000.jpg

Run general-purpose workloads on Google Axion Processors for exceptional performance, energy-efficiency and advanced capabilities.

Axion is but the latest in a long line of custom Google silicon. Since 2015 we’ve released five generations of Tensor Processing Units (TPU); in 2018 we released our first Video Coding Unit (VCU), achieving up to 33x more efficiency for video transcoding; in 2021, we doubled-down on custom compute by investing in “system on a chip” (SoC) designs, and released the first of three generations of Tensor chips for mobile devices. 

While our investments in compute accelerators have transformed our customers’ capabilities, general-purpose compute is and will remain a critical portion of our customers’ workloads. Analytics, information retrieval, and ML training and serving all require a huge amount of compute power. Customers and users who wish to maximize performance, reduce infrastructure costs, and meet sustainability goals have found that the rate of CPU improvements has slowed recently. Amdahl’s Law suggests that as accelerators continue to improve, general purpose compute will dominate the cost and limit the capability of our infrastructure unless we make commensurate investments to keep up.

Axion processors combine Google’s silicon expertise with Arm’s highest performing CPU cores to deliver instances with up to 30% better performance than the fastest general-purpose Arm-based instances available in the cloud today, up to 50% better performance and up to 60% better energy-efficiency than comparable current-generation x86-based instances1. That’s why we’ve already started deploying Google services like BigTable, Spanner, BigQuery, Blobstore, Pub/Sub, Google Earth Engine, and the YouTube Ads platform on current generation Arm-based servers and plan to deploy and scale these services and more on Axion soon.

Unrivaled performance and efficiency, underpinned by Titanium

Built using the Arm Neoverse V2 CPU, Axion processors deliver giant leaps in performance for general-purpose workloads like web and app servers, containerized microservices, open-source databases, in-memory caches, data analytics engines, media processing, CPU-based AI training and inferencing, and more. 

Axion is underpinned by Titanium, a system of purpose-built custom silicon microcontrollers and tiered scale-out offloads. Titanium offloads take care of platform operations like networking and security, so Axion processors have more capacity and improved performance for customer workloads. Titanium also offloads storage I/O processing to Hyperdisk, our new block storage service that decouples performance from instance size and that can be dynamically provisioned in real time. 

“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." - Rene Haas, CEO, Arm

Beyond performance, customers want to operate more efficiently and meet their sustainability goals. Google Cloud data centers are already 1.5X more efficient than the industry average and deliver 3X more computing power with the same amount of electrical power compared with five years ago2. We’ve set ambitious goals to operate our offices, campuses, and data centers on carbon-free energy, 24/7, and offer tools to help you report on carbon emissions. With Axion processors, customers can optimize for even more energy-efficiency. 

Axion - out-of-the-box application compatibility and interoperability

Google also has a rich history of contributions to the Arm ecosystem. We built and open sourced Android, Kubernetes, Tensorflow and the Go language, and worked closely with Arm and industry partners to optimize them for the Arm architecture. 

Axion is built on the standard Armv9 architecture and instruction set. Most recently, we contributed to SystemReady Virtual Environment (VE), Arm’s hardware and firmware interoperability standard that ensures common operating systems and software packages can seamlessly run on Arm-based servers and VMs, making it easier for customers to deploy Arm workloads on Google Cloud with limited-if-any code rewrites. Through this collaboration, we’re accessing an ecosystem of tens of thousands of cloud customers already deploying workloads and leveraging Arm-native software from hundreds of ISVs and open-source projects. 

Customers will be able to use Axion in many Google Cloud services including Google Compute Engine, Google Kubernetes Engine, Dataproc, Dataflow, Cloud Batch, and more.  Arm-compatible software and solutions are now available on the Google Cloud Marketplace, and we've recently launched preview support for Arm-based instances migration in the Migrate to Virtual Machines service.

What our customers and partners are saying

"We're thrilled to add application packages built on the new Axion Arm-based CPU for Google Cloud to the Bitnami by VMware Tanzu library. This will deliver significantly improved performance, attractive price-performance, and better sustainability for our users. We're excited to get our hands on the new Google Axion instances and do even more to help our customers streamline deployments and reduce their environmental footprint." - Mike Wookey, Senior Director R&D, Tanzu Division, Broadcom

"Organizations all over the world rely on CrowdStrike and our single platform, single agent architecture to stop cloud breaches. CrowdStrike delivers the industry’s best protection while being the fastest to deploy, so we’re excited about testing Google's new processor to discover power and efficiency gains." – Daniel Bernard, Chief Business Officer, CrowdStrike

“Our customers demand uncompromising cybersecurity protection that our systems provide. We're intrigued by the power and efficiency gains possible with the new Google Cloud's custom Arm-based CPU and plan to validate its capabilities as a way to provide even better threat detection and response capabilities to our customers.” - Tzach Segal, VP Business Development, Cybereason

“Datadog has been a trusted partner for customers adopting Arm-based virtual machines and an early adopter of Arm for our own operations. We’re excited about Google Cloud's announcement of the Axion processor and plan to evaluate it on our workloads as we help customers measure the cost and performance benefits of using Datadog on Google Cloud Arm instances.” - Yrieix Garnier, VP of Product Management, Datadog

“Elastic is committed to helping customers unlock the potential of all their structured and unstructured data efficiently at any scale. We look forward to testing Google Cloud's new custom Arm-based CPU and expect it to help us provide an even better Elastic customer experience on the Google Cloud Platform.” - Steve Kearns, VP of Product, Elastic 

“We’ve built a strong partnership with Google Cloud over many years and have seen the benefits of building on GCP Arm-based VMs. We can’t wait to see the remarkable improvements coming with the new generation Arm-based Axion processor.” - Joel Meyer, SVP, Engineering, OpenX  

"Snap empowers everyone to express themselves, live in the moment, learn about the world, and have fun together. We're constantly optimizing our infrastructure for performance and efficiency. Google's new Axion Arm-based CPU promises major leaps forward in both. The potential to serve our community with these gains while leading on our sustainability goals is incredibly exciting. We look forward to seeing the benefits of Axion-based virtual machines when they become available." - Korwin Smith, Sr Director of Engineering, Cloud Infrastructure, Snap

“WP Engine powers websites for more than 1.5 million customers across 150 countries. They rely on WP Engine to deliver on our core promises of performance, reliability, and security and we take that responsibility to heart. Our commitment to innovation and a customer-inspired mindset means we’re always looking for ways to further enhance our customers’ performance and confidence online. We’re excited to test Google’s new custom Arm-based processor and its anticipated performance and sustainability gains, and explore how they can empower our customers to create even more compelling websites and applications.” - Ramadass Prabhakar, SVP and CTO, WP Engine

Learn more

Virtual machines based on Axion processors will be available in preview in the coming months. Register your interest today! And if you’re here at Next ‘24, come learn more about Axion and other compute announcements in these related sessions:


1. Google Cloud Internal Data, 31 March 2024
2. Google Environmental Report, 2023, page 10.

Posted in