Google One-Ups AWS’ Second Billing Plan Google launched this week a per-second billing plan for renting virtual machines (VMs), just over a week after Amazon Web Services (AWS). The second billing cycle will begin Tuesday for VMs that are running on the following Google Cloud Platform services (GCP).

  • Compute Engine
  • Container Engine
  • Cloud Dataproc
  • App Engine

Google used to bill users by the minute for running VMs on its platform. This was the norm. According to Paul Nash, GCP Group Product Manager, per-second billing has been used for some GCP offerings such as GPUs, Persistent Disks and committed use discounts since their launch. Nash also pointed out that, unlike switching from a per hour to a a per minute billing structure, switching to per-minute billing will likely result in minimal savings for most organizations. He said that it was a fraction of one percent. Google’s announcement seems to be an answer to AWS last week’s move to introduce per-second billing to its cloud. Google’s billing changes seem to go beyond its rivals. Google, for instance, is implementing per second billing across a wide range of its services. However, AWS’ per second billing change only applies to Elastic Compute Cloud instances and Elastic Blockstore volumes. Google also stated that its per-second billing applies “all VMs”, including Linux and Windows. AWS’ billing scheme does not apply to Windows instances. Both platforms require a minimum of one minute for their respective per second billing plans.