Choose the Best: AWS vs. Azure vs. Google Cloud Comparison
Newcomers’ choice often tends to be driven by the costs of ownership and adoption when comparing different cloud providers. While price is an important factor, it should not be the decisive one. Each market player – Microsoft Azure, Amazon Web Services, and Google Cloud – come with important differences, making them more suitable for different companies and business tasks.
In fact, the cloud solutions offered by each company are largely influenced by their respective backgrounds. Amazon has been among the pioneers of aggregating, compiling and processing a large amount of structured and unstructured data. Google has a strong analytical methodology and Microsoft offerings are deeply rooted in the fact that they were among the first providers of computing services.
Different technological strengths and backgrounds have shaped the companies’ approach to delivering cloud services and additional functionality as this post will illustrate.
Top Cloud Services Providers: Market Overview
Amazon Web Services (AWS) currently dominates the IaaS/PaaS market with a 51% market share according to MarketWatch while Google with its rather young Compute Engine IaaS services gets the smallest share – 6%: :
Microsoft Azure usage, however, surged in the past year as the company has embraced the “cloud first” business strategy under the new CEO. As of lately, Azure became the primary choice for a SaaS cloud solution due to its extensive functionality and seamless integration options with other Microsoft products such as Visual Studio Tools for Azure.
Azure also is among the top choices for governments and now hosts some 6 million government cloud users. This fact can be largely attributed to increased security and the launch of special services catering towards this group of users – Microsoft Operations Management Suite and Azure Government Marketplace.
Google Cloud comes third in terms of market share, however, it is often identified as the fastest growing cloud services provider. G Suite alone now boasts 4 million paying customers, and Google has recently confirmed that its cloud services generate $1 billion per quarter.
To understand each solution better, it’s not enough to assess the company’s market share and user base though. We’ll further evaluate each provider based on different parameters including compute, storage, databases and pricing options available, along with the general overview of the core competitive features.
AWS vs. Azure vs. Google Cloud: Tech Characteristics Comparison
EC2 (Elastic Compute Cloud) – a solution that manages virtual machines. Comes with pre-configured settings that can be customized by the user.
CPU Limits: 1-40 CPU Memory Limits: 0.5-244 GB
Offers Virtual Machines that can be configured to manage, deploy and maintain OS and necessary server software.
CPU Limits: 1-32 CPU Memory Limits: 0.75-448 GB
GCE (Google Compute Engine) – functions similarly to AWS.
CPU Limits: 1 Shared-32 dedicated CPU Memory Limits: 0.6-208 GB
Amazon S3 (Simple Storage Service) comes with extensive documentation, proper community support and overall good reviews from users. Allows to physically ship the data to them in order to upload it to the cloud for you.
Costs: Starts at $0.023 per GB Per month
Azure Blob Storage is an elastic storage solution for large, unstructured data (up to 5 TB per item). It’s highly durable, easy-to-integrate with other applications and uses top-notch 256-bit AES encryption for data protection.
Costs: starts at $0.002 per GB for hot storage and goes down to $0.01 per GB per month for cool storage
Google Cloud Storage provides resumable data uploads; has great interoperability with other cloud storage services and provides strong read-after-write consistency for all upload operations.
Costs start at $0.007 – $0.014 per GB/month/
Amazon RDS (Relational Database Service) supports all major databases such as Oracle, PostgreSQL and others. The solution even handles patching, updating and some minor debugging.
DocumentDB that can handle high-performance databases.
Analytics & Big Data
Real-time data can be processed with Stream Analytics. The new Azure Data Lake is a hyper-scale data storage and analytics platform that enables to operationalize large data without the need to provision/manage computer clusters. Power BI is a neat dashboard and visualization tool. Cognitive Services suit enables advanced speech, vision, and natural language processing.
Cloud Dataproc is a fully managed Hadoop and Spark service. You can also use Cloud Dataflow to build custom data processing pipelines. Google has an advanced Machine Learning platform for training and hosting Tensorflow models. Additionally, APIs and pre-trained models are available for Speech, Voice, and Natural Language processing tasks.
Amazon Web Services (AWS)
In terms of PaaS and IaaS, AWS gives organizations extensive computing resources. It has excellent technical characteristics, and its particular strength lies in supporting a lot of users and operationalizing a lot of data. Large tech companies such as Netflix, Airbnb, Expedia and others use AWS EC2 to provide their services to people across the planet.
- Pricing plans are built based upon actual usage, rather than a set monthly fee. This is attractive to smaller organizations, especially before they scale.
- Storage is customizable, something that not all platforms provide, and the cost is based on amount and type of storage.
- Support fees are variable too and are tied to monthly usage. Again, this is a good feature for organizations that do not anticipate lots of support needs.
- AWS does offer lots of flexibility and customization and supports third-party integrations. It’s a good platform for hosting Linux.
- Scalability to support huge numbers of users is a plus – probably one of the reasons Netflix has chosen this platform.
- Boasts the highest number of industry standard compliance certifications including HIPAA, SOC 1/SSAE 16/ISAE 3402, SOC 2, SOC 3, PCI DSS Level 1, ISO 27001, FedRAMP, DIACAP and FISMA, ITAR, FIPS 140-2, CSA, MPAA.
- There is a significant learning curve with AWS.
- The pay-as-you-go features of AWS can bring costs up pretty quickly as an organization scales or as tech support needs increase.
- AWS does not include enterprise support by default. It should be purchased as an add-on service. The business tier support comes with up to a 10% premium on the customers overall AWS bill.
- Enterprise clients have to schedule a custom arrangement and negotiate all the contract terms, rather than opt-in online.
AWS probably holds the edge for organizations with web-scale applications that must support a lot of users. If you are looking for a platform that is feature-rich and highly scalable, once you become comfortable with it, AWS is a good choice. And, it continues to offer new features and updates in an attempt to attract more customers.
Microsoft Azure is a robust SaaS cloud solution as it offers attractive integrations with other Microsoft products, set up in a matter of minutes. Capable to support both Windows and Linux, it is also pre-configured with ready-to-deploy server applications and different languages.
Azure is a strong candidate both for smaller and larger companies, especially as a hosting solution for SAP systems .
- Azure is a much more user-friendly out-of-the-box solution than its competitors, with a straightforward setup process.
- Integration of Azure VM’s with other Microsoft products makes migration pretty smooth.
- Variety of pricing allows organizations of all sizes to choose this platform. And storage costs are fixed based upon amount.
- The other plus in terms of cost is that Microsoft rounds usage up to the nearest minute, while many others round up to the nearest hour.
- There are a variety of tech support plans, each with a fixed monthly cost, rather than variable pricing. Organizations know up front what they will be paying.
- Azure can offer a service level agreement of 99.95%, approximately 4.38 hours of downtime per year – not the kind of guarantee other providers deliver.
- All Azure services are subject to data transfer fees.
- You have to pay for a support account if you wish to place a ticket.
- The overall user experience with Azure ML components is very positive but they are reported to be somewhat slow.
- Data tends to be housed internationally. If you are bound by certain legal data hosting restrictions, you will have to verify the information with Microsoft.
Azure offers an easy, out-of-the-box getaway to cloud computing. It’s a great solution for organizations that want to migrate their VM’s to the cloud, yet do not at first anticipate needing lots of customization or rapid/large scaling.
Key Products & Features Google offers a full array of products/services to organizations of any size and needs. Among its provisions of computing, storage, networking, and big data / machine learning, it has options to appeal to a wide variety of organizations. Its biggest strength probably lies in data management, so if you have data-intensive applications, Google may be the best choice.
- Rapid Input/Output = less access time.
- Very strong in the areas of data analytics and storage segment.
- Seamless integration with other Google services.
- Sustained-use option does not require an upfront payment, such as Amazon RI’s.
- Includes the most advanced and robust ML services and products.
- Google positions its pricing model as “customer-friendly” and offers a free tier with some services included. Unlike other providers, the company also offers three types of discount packages that can help you slash over 80% off list prices if you are committed to staying long-term.
- Contract terms and special arrangements are more flexible compared to other providers.
- Most features are based only on Google technologies, so customers have no actual control over virtual machines.
- It’s difficult to migrate out of the platform if a change is desired.
- Still lacks decent server coverage in Europe and Asia, with no coverage in South America, which may be an issue for those wishing to store data locally.
- As a late entrant to IaaS market with its Google Compute Engine (GCE), the platform lacks some popular features already available with the competition.
Organizations that need data analytics and machine learning features will like Google Cloud. And any company looking to launch IoT devices/products will find this platform really attractive.
Choosing between different cloud providers is a serious business decision and the provided comparison only delivers a general overview of the key technical features, services and benefits. Before making the final call, it’s best to consult with the specialists that could assess your current business model and infrastructure; align the assessment with your ultimate goals and suggest product(s) that could help you attain those milestones.
Infopulse team has a successful track record of planning and implementing cloud migration projects for clients across different industries – ranging from the government to IT services providers. We’d be delighted to advise you on the right cloud solutions and help you bring your cloud strategy to life successfully.