Comparing Costs between Google Cloud and AWS
If you are comparing costs between Google Cloud and AWS, you will likely encounter the same issues as many people before you. Making a comparison between the two cloud platform providers is firstly complicated by Google and Amazon using different terminologies, and by Google´s seemingly random use of the terms “instance” and “virtual machine” to describe the same thing.
Thereafter, you have to account for factors such as Google´s per-minute billing (as opposed to Amazon´s per-hour billing), the options provided by Google to create your own custom machine types (as opposed to Amazon´s predefined compute sizing), and then networking, storage and load balancing. No wonder few people attempt comparing costs between Google Cloud and AWS twice!
Comparing Costs between Google Cloud and AWS for Compute Charges
To illustrate how difficult comparing costs between Google Cloud and AWS is, we are going to compare the prices the two cloud platform providers charge for their compute services. The table below shows the price per hour of running similar types of instances across different operating systems. The figures in bold indicate the cheapest per hour price, so it would appear the advantage is with AWS.
However, once you account for Google´s per-minute billing, a different story emerges. For example, the per-hour cost of running a two core general purpose m4.large instance on Linux is 10c per hour with AWS. The cost of running a similar n1-standard-2 instance is 10.4c per hour with Google. If you were to run the AWS instance for ninety minutes, the cost would be 20c because AWS charges by the hour. If you were to run the Google instance for ninety minutes, the cost would be 15.6c.
The difference in cost does not amount to much when comparing single instances against each other, but over hundreds of instances running for hundreds of hours, the difference can be substantial.
Taking Sustained Use Discounts into Account
If you leave instances running most the time, the per-minute/per-hour billing difference might not make such a great difference when comparing costs between Google Cloud and AWS. Nor might committing to a predetermined level of usage, as there is little to choose between Amazon´s “Reserved Instances” and Google´s “Committed Use” pricing plans. However, it is worth taking Google´s “Sustained Use” discounts into account when comparing costs between Google Cloud and AWS.
Sustained Use discounts are applied incrementally by Google depending on the percentage of each month instances are run. For example, an n1-standard-2 instance is charged at 10.4c per hour for the first 25% of the month, at 8.32c per hour for the second 25% of the month, at 6.24c per hour for the third 25% of the month, and at 4.16c per hour for the final 25% of the month. Over a thirty day month the cost of running the instance would be $52.40 rather than $74.88 – a saving of 30%.
Most businesses do not run all their instances all the time, so Google has made it possible for businesses to combine non-overlapping instances in order to maximize the monthly discount. This process is known as “inferred instances”. The great thing about Sustained Use discounts and inferred instances is there are neither commitments nor prepayments required in order to take advantage of them. Certainly something to take into account when comparing costs between Google Cloud and AWS.
Maximize Sustained Use Discounts with ParkMyCloud
By using ParkMyCloud to automate stop/start times for non-production instances, it is possible to increase Sustained Use discounts beyond 30%. For example, if a business were to “park” its non-production instances outside of typical 8:00 a.m. to 8:00 p.m. working hours, Google would only charge 264 hours of compute time (assuming a thirty day month with four weekends).
Based on an n1-standard-2 instance, the first 180 hours of the month would be charged at 10.4c and the next 84 hours charged at 8.32c for a total of $25.71 – a saving of 51% on the maximum Sustained Use discount. If you are still considering deploying your instances on AWS after comparing costs between Google Cloud and AWS, the savings achievable with ParkMyCloud are even more substantial.
To find how much you could save by using ParkMyCloud to maximize Sustained Use discounts, you are invited to try our scheduling software for free. To take advantage of this opportunity, click the “Try it Free” button or speak with a member of our friendly Customer Service Team. We will be happy to answer any questions about parking non-production instances on Google Cloud or AWS, and how ParkMyCloud can help generate further savings by improving the governance of your cloud accounts.