This content is part of the Essential Guide: An insider's look at AWS re:Invent 2014
Evaluate Weigh the pros and cons of technologies, products and projects you are considering.

AWS, Google Compute Engine or both?

AWS and Google are noted leaders in cloud. While AWS stands out on some features, Google holds its own in others. That doesn't mean you have to choose between them.

The market recognizes AWS and Google as the leaders in cloud computing. However, because of its range of cloud capabilities up the stack, AWS is currently leading the race. But that doesn't mean it's the right choice for your company. One way to decide is to compare both clouds on price, global reach and capacity. And even then, it may be difficult to settle on AWS, Google or both.

The price/performance debate: Industry experts and research show a balance between Amazon Web Services (AWS) and Google Compute Engine (GCE). Google's price per hour for the same instance specifications (more or less) is lower and it provides better performance results compared to AWS.

A key difference on GCE's price model is that the cloud provider charges by the minute, while AWS charges by the hour. But there are a few ways to cut costs with AWS, such as using Reserved Instances and spot instances for projects like research and development and non-mission-critical applications.

Global footprint and capacity: AWS is a more mature cloud provider, with 11 infrastructure regions, including GovCloud and the new region in Germany. Google only has three regions. AWS also has a much larger capacity and a wider variety of instance families than GCE. AWS offers up to 1 TB per disk; GCE provides 10 TB of storage per disk.

When it comes to capacity, GCE allows you to mount a single disk to several virtual machines (VMs) in read mode, which is  important in distributed environments. Google also offers no-warm-up and quick response capabilities for load balancing. But travel up the stack a bit and AWS offers many more data and storage capabilities, such as Elastic MapReduce, Hadoop and Redshift.

Cloud payers and players: Another important factor to compare is the cloud ecosystem and marketplace. This is an area where Google lags behind AWS.

Amazon has thousands of partners, according to Andy Jassy, senior vice president for Web services, during his AWS re:Invent 2014 keynote. The AWS ecosystem is huge and growing rapidly with partners, consultants and managed service providers. And even though the AWS Marketplace had a slow start last year, it is now gaining traction. Consumers can provision various services and spin up packaged, out-of-the-box applications from the AWS Marketplace with the click of the mouse.

Making the not-so-clear choice

So which of these two major cloud vendors should companies choose? The simple answer is both. It no longer needs to be a choice of one cloud over the other.
But building your environment with the ability to move between cloud services is no easy feat. To be truly cloud agnostic, companies must go with the common denominator features of both AWS and Google to normalize the environment.

And the AWS re:Invent expo hall is filled with a slew of vendors, such as Ravello Systems, that have developed this normalization layer, providing seamless portability and interoperability of the complete application stack. These tools can support on-premises migrations to the cloud and between clouds, giving the market a small break in the mounting AWS, Google argument.  

About the author:
Ofir Nachmani is a business technology advisor, blogger and lecturer. Ofir's extensive experience in the world of business technology has made his critically acclaimed blog,, the go-to guide for modern technology startups and developers in the world of cloud computing. Today he advises organizations, leading them through new IT market modifications, while building and executing a modern go-to-market strategy.

Dig Deeper on Amazon EC2 (Elastic Compute Cloud) management

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.

While there is becoming more parity between the large cloud providers, most decisions still boil down to on-prem vs. off-prem (or a mix of both), not so much multi-cloud. The cloud migration technologies are still somewhat immature, although things like Ravello Systems, Dell Multi-Cloud Manager (enStratius), Clikr and a few others are helping this. 
Good points Brian. To migrate apps to the cloud, is a great choice.
The “cloud” is different than its predecessors. While the concept of having someone like Amazon or Google provide the computing power to support your application is old hat, the technology underlying the delivery of these services is much more advanced and complex. We cannot determine the right choice based only on the price sheet or cost of running the application. We need the ability to see into the cloud, meter the performance being delivered and know the true cost and quality of the service being delivered. There is research that shows the same machines provisioned on the top three cloud service providers delivers different performance for the identical workload. Moreover, the variance occurs within providers! Without very advanced monitoring and control, we can easily be fooled into thinking we are getting the best value. A new discipline, called Platform Performance Management, will become increasingly important when choosing or operating cloud hosted and internally hosted applications. The only player I know in this space today is Krystallize Technologies.