AWS vs Azure vs Google Cloud: A Quick Comparison

AWS vs Azure vs Google Cloud: A Quick Comparison

The competition for dominating the cloud world has become a fierce race between the three incredible and leading market leaders – AWS, Azure, and Google Cloud. They keep adding features, modifying their prices, and try different strategies to win that “cloud throne.” 

However, each has its own set of perks and drawbacks, depending on the use case. If you are just starting out on these technologies, you will have confusions for sure. So, here is a quick comparison of AWS vs Azure vs Google Cloud to help you understand which is better and why? 

Before we explain and give you a comparison between these three vendors, let’s start with some statistics.

AWS:

Amazon Web Services (AWS) reported a total sales of 7.7 billion USD in the first quarter of 2019, which was 5.44 billion USD same time in the previous year. The revenue of AWS inclined by 41% when compared to the last year’s first quarter. (Source)

Amazon service

Image Credit: Geekwire

Microsoft Azure:

While Amazon has a clear, separate revenue pouring in from AWS, Microsoft has a hazier “commercial cloud business” – because it not just includes Azure, but Dynamics 365, Office 365, and other components of the Productivity & Business Process Division. Though this might be frustrating to many users, it simply cannot be compared to AWS directly. 

Microsoft reported that its commercial cloud business saw a growth rate of 41% in Q1 of 2019 at 9.6 billion USD. However, Microsoft didn’t give out much details in its announcement. (Source)

It just gave out one revenue number associated with Azure: 73%. Well, that’s their overall growth rate. Microsoft has buried its Azure-related revenues in a bigger pile of "Commercial Cloud."

Everyone knows that Azure grew at 98% a year back. Since then, its growth rate is going down every quarter. But this is totally normal. When revenue numbers go high, sustaining the same progress is usually difficult.


Image Source: Geekwire

Google Cloud:

Like Microsoft, even Google played the same game. It did not report specific numbers on its cloud revenue once again.

Google’s parent company - Alphabet Inc, reported a total revenue of 36.34 billion USD in Q1 2019, which increased by 17% from 31.15 billion USD for the same quarter last year. 

The revenue of Google Cloud Platform (GCP) is included in its “other” revenue category, along with Google Play, G-Suite, and hardware like Nest. This category reported a total revenue of 5.45 billion USD for Q1 2019, up by 25% in the same time last year, which was at 4.25 USD billion.

The Market Share: 2018 vs. 2019

According to Canalys, a renowned analyst company, reported that AWS was in the lead in 2018, owning about 1/3rd of the cloud market share, Microsoft in the second at 15%, and Google in the third at 5%. 

In 2019, Canalys reported a total growth rate of 42% in the cloud business. Though AWS had the largest sales gain at $2.3 billion (Year-On-Year increase), Canalys reports Google Cloud & Azure with greater percentage growths.


Image Source: Canalys

AWS vs Azure vs Google Cloud Market Share: The Winner

Based on the recent reports, it is apparent that AWS is in the lead. 

Bezos said:

“AWS had the unusual advantage of a 7-year head start before facing like-minded competition. As a result, the AWS services are by far the most evolved and most functionality-rich.”

The market shares of three leading cloud service providers are at:

•    62% - Amazon Web Services (AWS)
•    20% - Microsoft Azure
•    12% - Google Cloud Platform (GCP)

AWS remains in the lead as of now. Having said that, it will be exciting to see how its competitors play out. Perhaps, next year by this time, the revenue numbers would have broken out and we will be able to say it for sure.

AWS vs Azure vs Google Cloud: A Quick Comparison 

Choosing the best CSP (Cloud Service Provider) is a hard decision. The scenario isn’t about what option you want to go with but rather how can your business accomplish optimal performance and further, distribute risks across various vendors—while covering the cloud’s storage & compute cost at the same time.

According to a recent survey conducted by Forrester Consulting in July 2018 from an enterprise-class cloud firm called Virtustream said,

“Multi-cloud arises from Changing Cloud Priorities.” 

The survey involved 727 well-known cloud technology decision makers with businesses involving over 1000 employees. The report outlined how varying business significances are driving companies to implement multi-cloud strategies. 

About 86% (majority) of the respondents described their present cloud strategy to be multi-cloud in their respective organizations. Moreover, the study found that 60% of the enterprises are moving or already have moved their mission-critical apps to the public cloud.

As far as the survey reports are considered, we can learn that adopting multi-cloud strategies can:

•    Save cost 
•    Improve performance
•    Enhance delivery times

 Service-to-Service Comparison

Typically, enterprises look to cloud service providers for 3 service levels: 

•    IaaS (Infrastructure as a Service) – the “storage & computing” capacity
•    PaaS (Platform as a Service) – the entire environments to develop, deploy, and manage the web apps)
•    SaaS (Software as a Service) – the performance and secured hosting for the applications.

Note: We are not going to compare pricing of these services, as it is complicated to achieve apples-to-apples kind of comparisons without a clear use case. After you determine your business’s CSP requirements, calculate the price of each CSP to check if there is any significant cost difference. This is the ideal way to choose a vendor. 

1. Storage

AWS, Azure, and GCP – all offer an extensive range of objects, file storage, and blocks for use cases involving both primary as well as secondary storage. 

While object storage is great for handling huge amounts of unstructured data (videos, images, etc.), block storage offers enhanced performance to structured data (transactional). 

The storage tiers offer various accessibility & latency levels to meet the requirements of both inactive (cold) and active (hot) data cost-effectively. 
 

  • For managing DR (Disaster Recovery) & Backup Services, Azure wins it.
  • For managing your hybrid architectures, Azure & AWS have inbuilt services whereas GCP depends on its partners.

 

  AWS Azure Google Cloud
VM Disk Storage Amazon EBS (Elastic Block Store) Azure Managed Disk Persistent Disks (both SSD & HDD)
Object Storage Amazon S3 (Simple Storage Services): AWS’s first public service Blob Storages Google’s Cloud Storage
Disaster Recovery Offers a few disaster recovery services (cloud-based) DRaaS (Site Recovery) Doesn’t provide any disaster recovery service
Backup Recovery Often, Amazon S3 is used as secondary backup storage The backup facility is inbuilt in the Azure platform Doesn’t provide any backup service
Transfer of Bulk Data ~ AWS Import & Export
~ AWS SnowMobile
~ AWS Snowball – based on the device
 
~ Azure Import & Export
~ Azure’s Data Box Disk Service
Transfer service for storage
Archive Storage ~ S3 one-zone infrequent access
~ Amazon Glacier – has data querying abilities 
Azure’s long-term storage:
~ Archive Storage (offers blob storage offline)
~ Cool Blob Storage (less availability when compared to “Hot”)
 

Archival cloud storage:

~ Coldline (offers the lowest frequency)
~ Nearline (offers low frequency)

Hybrid Support

AWS Storage Gateway:

Offers a virtual tape, managed infrastructure in a hybrid environment
 

StorSimple

Offers cloud storage for hybrid at an enterprise-grade 

Depends on its partner - Egnyte
 

2. Compute

AWS, Azure, and GCP – all offer a comprehensive range of predefined instances, which define every virtual server that is launched, RAM, CPU (or GPU) processor type, the number of vGPU/vCPU cores, and local storage (temporary). 

The predefined instance type will determine:

•    Compute speed
•    I/O speed 
•    Performance parameters

It lets you to optimize performance/price based on the workload requirement. 
These cloud service providers offer pay-as-you-go options to handle the scaling, deployment, and balancing of web apps automatically, given that they are built in leading frameworks like Node.js, Java, Python, PHP, Ruby, etc. 

 

  • AWS provides auto-scaling option without any extra charge. However, it also depends on the scaling plan you define.
  • Azure provides auto-scaling option for every app or a set of apps/virtual machines.
  • Google Cloud provides auto-scaling option only in “Managed Instance Group” platform.

 

Both Azure & AWS offer services, which lets you to create a VPS (Virtual Private Server) in some clicks. However, GCP doesn’t offer this proficiency yet.

 
  AWS Azure Google Cloud
PaaS Amazon’s Elastic Beanstalk Azure’s Cloud Services GAE (Google App Engine)
Virtual servers Amazon EC2 (Elastic Compute Cloud) Virtual Machine for Windows/Linux server Google’s Compute Engine
Support for VPS (Virtual Private Server)  Amazon Lightsail VM (Virtual Machine) Image Not Applicable
Scaling Amazon’s Auto-scaling Azure’s Auto-scale
Scale sets of virtual machines (high-availability and hyperscale apps) 
Via instance groups
 


3. Management Tools

Managing & orchestrating your cloud resources across diverse business units as well as other complex infrastructures is a challenge. 

Happily, all these three cloud platforms provide visibility and helps in streamlining your organizational processes. Their services range from catalogues of approved services to predefined deployment templates and centralized access control. 

 
  AWS Azure Google Cloud
Templates for cloud deployment

Amazon’s CloudFormation: 

~ Provisions cloud resources.
~ Text files for modelling 

Azure’s Resource Manager: 

~ Deploy and control access to resources that are categorized; includes templates 

Google Resource Manager: 

~ Categorize, organize, and controlled access to resources; manage & track projects 
Cloud Deployment Manager: 

~ Features template- driven deployment

Server manage- ment services

AWS’s System Manager:

~ Automation and visibility across resources

Azure’s Operational Insights: 

~ SaaS; Operational data analysis

Not applicable
Server automation

~ Amazon OpsWorks

~ Amazon Service Catalog

~ Azure Resource Manager 
~ VM extensions: Post-deployment configuration and automation
~ Azure Automation
Not Applicable
Logging and Monitoring

Amazon’s CloudWatch:

~ Visibility of apps in real-time 

AWS’s CloudTrail: 
~ Logging and monitoring AWS accounts

MS Azure Monitor, with Log Analytics & Application Insights  Google Cloud StackDriver, which includes logging, monitoring, tracing, error reporting, and debugging
 

Conclusion

Rather than going with the “best,” identify how will you distribute your workloads optimally across various CSPs. When you try to implement your multi-cloud strategies, remember that the principal categories of compute, management tools, and storage in Azure & AWS offer a mature and comprehensive stack when compared to GCP.

Generally, the products and services offered by AWS are the most wide-ranging, but they are equally challenging to manage and navigate. Also see if your organization is using Microsoft’s development tools, Office productivity apps, or Windows servers. If yes, then you will find Azure pretty straightforward to integrate.