Google touts lower costs with new family of general purpose VMs

Now available in beta, with new regions in the works

Comments

Google has launched a new family of general-purpose virtual machine (VM) instances for Google Compute Engine that the company claims offers the lowest total cost of ownership of any VM in Google Cloud.

Google Cloud’s new E2 VM instances, which are now available in beta, make use of dynamic resource management capabilities to deliver sustained performance, flexible configurations and, according to a blog post by Google Compute head of products June Yang, the “best total cost of ownership of any of our VMs”.

Moreover, the E2 VMs are designed to make better use of hardware resources by using resource balancing technologies developed for Google’s own latency-critical services -- another factor the vendor claims drives costs for users of the new VMs.

As a result, according to Yang, the E2 VMs offer similar performance to N1 configurations, which it claims are comparable, but represent 31 per cent savings compared to N1 machine types.

Based on x86 chips from Intel and AMD, the new family of virtual machines are designed to be used for a broad range of workloads, including web servers, business-critical applications, small-to-medium sized databases and development environments.

E2 instance can be used with up to 16 vCPUs and 128 GB of memory, while users can get away with provisioning only the resources that they need with 15 new predefined configurations, according to Google.

While the new E2 VMs are in the process of being rolled out in just eight regions -- Iowa, South Carolina, Oregon, Northern Virginia, Belgium, Netherlands, Taiwan and Singapore -- more regions are in the works.

As at launch, Yang said that Google is offering E2 machine types as custom VM shapes or predefined configurations. Full details below:

Yang said the company is also introducing new shared-core instances, similar to the company’s f1-micro and g1-small machine types.

“These are a great fit for smaller workloads like micro-services or development environments that don’t require the full vCPU,” Yang said in her post. Below are the details of the new shared-core instances: