Google Compute Engine vs. Amazon Web Services

The cloud battle is on.

google amazon
A new era of cloud battles

This week Google Compute Engine became generally available, meaning that anyone can sign up to use the same on-demand virtual machines and storage that the Internet’s leading search website uses to power its own operations. In light of this news, Gartner analyst Lydia Leong says the cloud computing market has entered a new era. Amazon Web Services has been the kingpin of the market, but it now has a more legitimate competitor than ever. Here’s a look at how the two cloud offerings compare.

Google Compute Engine vs. Amazon Web Services cloud: The battle is on

Virtual Machines: Google
Virtual Machines: Google

On-demand virtual machines are a good measure of a cloud offering. Google has 15 virtual machine sizes to choose from across four categories: standard, high memory, high CPU and shared core VMs. They’re all Linux-based and natively support Red Hat Enterprise Linux and SUSE Linux, but can work with any Linux distribution. Unlike with AWS, prices are pro-rated by the minute and not charged exclusively by the hour. Prices start at $0.14 per hour for a single core 3.75GB of memory standard disk. 

Virtual Machines: AWS
Virtual Machines: AWS

AWS has a much broader set of virtual machines available to customers. Between its on-demand virtual machines, reserved instance VMs and spot market VMs, there are dozens of options. There are general purpose VMs, high CPU, high graphic output VMs, storage and memory optimized VMs, along with micro and macro VM sizes, among others. AWS offers not only Linux VMs, but Microsoft Windows ones too. Its m1.medium VM, which comes with 3.75GB of memory, starts at $0.12 per hour. All VMs are priced by the hour.

Storage: Google
Storage: Google

GCE comes with persistent disk storage, which means that the storage associated with virtual machines can live beyond the life of the VM and can be connected to multiple different VMs. Google also has three main database storage platforms, including a MySQL relational database named Cloud SQL; Cloud Storage, which is a highly-available object storage; and Cloud Datastore, which is a managed NoSQL database. Cloud Datastore is still in preview mode.

Storage and databases: AWS
Storage and databases: AWS

AWS has a variety of storage options, including Simple Storage Service (S3); Elastic Block Storage (EBS), which is a block storage service; Archival storage from Amazon Glacier - which provides lower-cost storage with longer retrieval times; and Cloud Storage Gateway, which is software that provides access to frequently used cached data in Amazon’s cloud. It has a wide variety of databases too, including AWS Relational Database Service (RDS); DynamoDB, its NoSQL database; and RedShift, the petabyte-scale data warehouse; as well as other databases from partners.

Analytics
Analytics

AWS has Elastic Map Reduce, a Hadoop-like computing engine for data analytics. It just released Amazon Kinesis as well, which is a real-time processing analytics engine. Google, meanwhile has BigQuery, which is its own Hadoop-like analytics platform.

Partners
Partners

This is an area where AWS has a clear advantage over Google. Google lists 22 partners, including cloud integration and performance optimization tools for GCE. AWS, by contrast has a list of 22 categories of partners that have optimized their products and services to run on top of its infrastructure, including big-name enterprise apps from SAP and Oracle.

More information on AWS Marketplace here

More information on Google cloud partner here

Security and compliance
Security and compliance

Both companies stress the security of their cloud platforms. Google says that it stores data at rest in encrypted fashion and has certifications including ISO 27001 and SSAE-16. AWS, has an entire region of its cloud dedicated to government workloads, and its cloud can be enabled to be healthcare and payment card industry compliant.

Other notable features: Google
Other notable features: Google

Google’s other major cloud offering is Google App Engine, its application development platform as a service (PaaS). This was the first cloud platform Google released, but it’s still in preview mode. So Google Compute Engine made it to general availability faster even though it debuted later. Google also has Cloud Endpoints, which are API tools for building applications that use Google cloud-based resources.

Other notable features: AWS
Other notable features: AWS

AWS has a bevy of ancillary features in its cloud, ranging from auto-scaling and load balancing (Google says this is included automatically in GCE), along with AWS Virtual Private Cloud, and AWS CloudFront, which is its growing content delivery network. Route 53 is a domain name system; Beanstalk is a tool for more easily managing and deploying applications in Amazon’s cloud. There are many more too, which can be found here.

New features: Google
New features: Google

With the general availability of GCE, Google rolled out new support for Linux OSs, including RHEL and SUSE Linux, as well as new 16-core instances for high-CPU workloads. Google also reduced its prices by about 10% and rolled out the live migration feature. Pictured is a photo, courtesy of Google, of one of its data centers.

New features: AWS
New features: AWS

At its re:Invent event earlier this year, AWS released a host of products and services including Amazon Workspaces, a cloud-based virtual desktop service, along with new graphics processing instance types, the Kinesis data processing application and support for open source PostgreSQL. AWS says it has added more than 250 features to its cloud this year alone.

microsoft ibm rackspace
Don’t forget about us!

The entrance of GCE into the IaaS market marks a new competitive dynamic, but it’s already a crowded market. Other competitors like Microsoft with its Azure offering, IBM with SoftLayer, and Rackspace with its offerings, along with other companies like Joyent and VMware, are each attempting to claim a piece of the cloud computing IaaS pie.