What took them so long! In what is clearly a reaction to the move AWS pulled about a week ago, Google has announced per-second billing for its cloud, the Google Compute Engine.
This is a very similar move to what Amazon Web Services announced earlier this month.
Like AWS, Google also charges for a minimum of one minute.
However, unlike AWS users that have to wait until October 2, customers using the Google cloud can start immediately. And it also one-ups Amazon in extending this to virtual machines that run premium operating systems like Windows Server, Red Hat Enterprise Linux, and SUSE Enterprise Linux Server.
Amazon Web Services, meanwhile, only offers per-second billing for basic Linux instances, and not for Windows Server and other Linux distributions on its platform.
All of which still feature a separate hourly charge.
That said, as Paul Nash, Group Product Manager for Compute Engine argues, per-second billing only results in very small billing changes, and that too, for applications where scaling up and down quickly makes sense:
“This is probably why we haven’t heard many customers asking for per-second. But, we don’t want to make you choose between your morning coffee and your core hours, so we’re pleased to bring per-second billing to your VMs, with a one-minute minimum.”
Examples of said applications being websites that receive a sudden burst of traffic, mobile apps and games that see heavy uses during certain times of the day, and data processing jobs.
Frankly, this is very much the case with AWS too.
Anyway, now that Google has made the move, we can probably look forward to Microsoft following suit in the very near future. Azure already offers per-second billing for containers, but then again, so did Google Compute Engine prior to today, for its Persistent Disks, GPUs, and committed use cases.
You can check out the details of this latest pricing change here.