Google has announced a new type of virtual machine (VM) that has “dynamic resource management” functions, which, in many cases, can drastically reduce the total cost of ownership.
The new VMs, called the E2 family, are intended for workloads that do not require a very large instance type, access to graphical processors or local solid state storage. That’s what June Yang, director of product management at Google Cloud, wrote in a blog post.
According to Google, the E2-VMs offer cost savings because they run on a custom central processing unit scheduler. This means a central processor that dynamically maps the virtual CPU and memory power of a physical CPU and memory. This ensures that the VMs are more efficient, and therefore customers are only billed for the computing power they actually use.
Comparable performance, lower costs
Yang further reports that the dynamic resource management features ensure that the E2-VM’s performance is comparable to that of Google’s N1 family, but compared to average savings of 31 percent on the total cost of ownership. The new VMs are also very flexible, as they have 15 pre-defined configurations. These configurations range from just 2 virtual CPUs and 2GB of memory to 16 CPUs and 128 gigabytes of memory.
“Your VMs get reliable and sustained performance at a consistent low price point,” Yang stated. “Unlike comparable options from other cloud providers, E2 VMs can sustain high CPU load without artificial throttling or complicated pricing.”
Shared Core Instances
Furthermore, Google has revealed a number of shared-core instances similar to Google’s f1 and g1 series. These instances also come with dynamic resource management, which should provide a good way to run small workloads such as microservices and development environments that don’t necessarily need a full vCPU.