Although it’s been a long wait, Google Cloud announced its first Arm-based virtual machines today.
This follows AWS with its Graviton instances and Azure which recently launched Arm virtual machines.
Google Cloud follows Azure’s lead and uses chips from Ampere, whereas AWS made its own chips.
These new VMs are currently in preview and will be joining Google Cloud’s Tau VMs under the “Tau T2A” moniker.
This line was launched nearly one year ago using AMD Milan processors to provide a better price/performance ratio.
“We are thrilled to expand the rich choices that we offer with Intel and AMD and to enter the Arm ecosystem to give our customers more choice and flexibility.
In a press conference ahead of today’s announcement, Sachin Gupta (VP and GM of infrastructure at Google Cloud) stated that the company supports a wide range of operating systems, databases and programming languages.
These new chips will be available in predefined SKUs that can hold up to 48 vCPUs and up to 4GB of memory.
These VMs can support up to 32 Gbps network bandwidth and the standard range of storage options in the Google Cloud ecosystem.
Google claims that these CPU specs make the machines suitable for many workloads, including web servers, containerized microservices and data-logging applications.
Google views these as its cost-performance optimized solutions, similar to the AMD-powered Tau chips. For example, a 32-core Tau T2A virtual machine in Google Cloud’s uscentral1 region will run $1.232 an hour.
These machines will allow users to run RHEL and CentOS on them, as well as Google’s Container-Optimized OS to run containerized apps.
Popular Stories You Shouldn’t Miss:
Arm support is now a standard feature for many OS and software vendors. This will greatly increase the utility of these VMs (and those of Google’s rivals).
The new VMs will soon be available at other data centres.
Jeff Wittich, Chief Products Officer at Ampere Computing, stated that “Ampere Altra Cloud Native Processors are designed from the ground-up to meet the requirements of modern cloud applications.”
“Our close collaboration has led to the launch of new, price-performance optimized Tau T2A instances. These enable scale-out applications to deploy quickly and efficiently.
These VMs can be used as part of Google Cloud’s Compute Engine. Google also supports them as part of its Kubernetes Engine.
This includes the Dataflow stream, batch processing service, and Batch. Google has also launched a fully managed job scheduler to manage batch jobs.
Gupta stated that this new capability will benefit major use cases in throughput-oriented computing, such as electronic design automation and weather forecasting.
This new service offers unprecedented flexibility in the location, cost and time of cloud capacity for batch jobs.
Popular Stories You Shouldn’t Miss: