When we first heard that Google Cloud Platform was coming to Australia, we cycled through those responses and started working with the platform to understand the benefits and constraints of that platform and how that compares to AWS. Now that Google Cloud Platform has launched here, we’re going to share those insights in a series of blog posts, starting with this one:
Now generally we wouldn’t rush straight to pricing when talking about cloud. We’re cloud people so we understand that cost isn’t always a straightforward topic and it’s certainly not the only benefit of using cloud. However, if you’re already familiar with AWS pricing then you’ll probably be interested in the cost savings you’ll get from moving to Google Cloud Platform. As Gartner noted in the 2017 Infrastructure-as-a-Service magic quadrant analysis - AWS pricing structure can be complex.
Understanding the points of difference in Google’s pricing approach will help you to identify potential monthly cloud cost savings, which I would argue should include reducing the overhead required to plan and manage your cloud financial operations. This can feed into a business case for moving some workloads to Google Cloud Platform.
Google are 30-60% cheaper than their competitors for standard workloads. That’s a pretty big call so let’s break down what that actually means.
1. Google Cloud Platform has a lower rack rate than their competitors, so the base rate of services is already discounted.
2. A loyalty discount referred to as ‘sustained usage discounts’ is automatically applied to your monthly bill if you continue to run a virtual machine for more than 25% of the days in a month, up to a maximum of 30% discount for continuous running throughout the month.
3. You are charged for use by the minute, rather than by the hour. If you’re running Dynamic workloads such as Hadoop, clusters may be spun up for only 30 minutes per day- and you only pay for 30 minutes, not an hour. This effectively halves the cost of running those workloads in cloud.
Those are the standard pricing features that are automatically granted when you’re using Google Cloud Platform. There are of course additional levers that you can control to further optimise pricing. The table below outlines the familiar pricing concepts of AWS and introduces the equivalent Google Cloud Platform offerings.
Amazon Web Services
Google Cloud Platform
Hourly billing: On AWS billing is hourly with any partial hours rolled up to the next hour. So if a job runs for 30 minutes you will be charged for one hour.
Per minute billing: On Google Cloud Platform, you pay per minute, not per hour.
Spot pricing: Amazon EC2 Spot instances allow you to bid on spare Amazon EC2 computing capacity.
Preemptible VM instances: This is the same concept as AWS Spot pricing but Preemptible VM instances have a set price, not a market (bid) price, making them easier to budget for. Customers save on average 80% of the standard price of VMs.
Reserved instances: Get a discount on VMs by locking into a 1-3 year agreement. Reserved Instances are available in 3 options – All up-front, partial up-front, or no upfront payments (NURI).
Committed use discounts: Get a discount on VMs over a 1-3 year agreement without upfront fees or instance-type lockin. Customers will be able to change the types of VMs they’re running, which is a key differentiator for Google Cloud Platform over AWS Reserved instances.
Trusted advisor: Performance analysis is performed by the Trusted Advisor service. This is available as part of business or enterprise support at a cost of ~$1000-$15000k/month
Rightsizing recommendations: Rightsizing is FREE service that analyzes performance of your instances and proactively recommends the optimal number of cores and memory based on the instance's usage over time.
Sustained Use Discounts: Automatically lower the price of your virtual machines (VMs) when you use them to run sustained workloads.
Custom machine types: You can customise virtual machine types to your needs using customisation sliders. Create a machine type with as little as 1 vCPU and up to 64 vCPUs, or any even number of vCPUs in between. Memory can be configured up to 6.5 GB of RAM per vCPU. By customising your machine type, rather than oversizing, customer discount implications up to 50% off.
How can Google have such competitive pricing?
Google are able to pass data centre savings onto customers because they engineer and optimise the entire platform end-to-end. Google operates at such a scale that they have built their own proprietary systems, including purpose-built chips, protocols, servers, storage, and switches.
In true Google style, Deep Mind’s artificial intelligence has been applied to the problem of optimising power usage in Google data centers. As a result they were able to reduce electricity required to cool data centres by up to 40% per month. On average, a Google data center uses 50% less energy than a typical data center. Awesome savings for customers and the environment.
Calculating the cost of moving
It probably cost you a lot to get into a cloud environment. And you may now be thinking that is a sunk cost and it will be equally expensive to move to another provider. If you’ve ‘lifted and shifted’ virtual machines into AWS EC2 then the news is good. Google have worked with CloudEndure to provide a free VM migration tool that automates migration from on-premise, or other cloud providers, to GCP. This tool is self-service and if you’d like the support of people who’ve done this before then there’s a whole ecosystem of partners, such as Axalon, available to help.
So you're interested in finding out more?
You can get started by signing up (use the $300 Free Tier button above)...
When you’re ready, we can help you on your Google Cloud Journey.
Any questions? Just ask.
Axalon help our customers to operate in real-time and to harness Google’s world leading capability in cloud continuous delivery, machine learning and analytics to build new business models.
We are Google Cloud Partners who also have extensive experience with Amazon Web Services. Axalon are advocates of open source and interoperable cloud. We are focused on building capability in the technology community through sponsoring the Google Developer Group - Cloud Chapter, and running training courses on the Google Cloud Platform, Kubernetes, DevOps and Machine Learning.
Our customers are busy enterprises and governments competing for customer engagement and market share in a rapidly changing environment. They have a vision but may be constrained by their current technology, processes or capability.
Check out our products and services at Axalon.io. If you found this blog post useful please subscribe below for future communications and/or ‘share’ this post.