Table of Contents
Cloud computing has revolutionized the way organizations access computing resources and build applications. As the two leading public cloud providers today, Azure and AWS offer excellent scalability, flexibility and cost efficiency. But they differ in their service offerings, pricing, cloud architectures and overall philosophies.
Introduction
Launched in 2006, Amazon Web Services (AWS) pioneered the Infrastructure-as-a-Service (IaaS) model, renting cloud computing, storage and networking services to organizations on a pay-as-you-go basis. This removed the need for enterprises to invest in hardware upfront. AWS has been the dominant IaaS provider over the past decade.
Microsoft Azure entered the scene in 2010 as a competitor to AWS. While AWS leads in market share with 32%, Azure has established itself firmly in second place with a 17% market share in the $182 billion cloud infrastructure market [1].
Both platforms now offer a vast array of services including:
- Compute – Virtual machines, auto-scaling, containers, serverless computing
- Storage & Content Delivery – Object storage, block storage, archive storage, CDN
- Databases – Relational databases, NoSQL databases, data warehouses, caches
- Networking – VPN, load balancing, traffic management, hybrid connectivity
- Analytics – Business intelligence, real-time analytics, big data
- Security & Identity – Role-based access, password management, key management
- Application Services – App hosting, notification engines, workflow engines
- Management Tools – Resource management, monitoring, cloud automation
However, there are some key differences between the two platforms that are important to understand when making a decision. This article provides an in-depth, independent comparison across some of the most important decision factors.
Feature Comparison
Compute Power
AWS offers a very wide range of instance types optimized for different workloads such as general purpose, compute optimized, memory optimized, accelerated computing, and storage optimized. The options go from nano instances with 1 vCPU & 0.5 GiB RAM all the way up to x1e.32xlarge with 128 vCPU & 3,904 GiB RAM.
Azure also provides flexible virtual machine sizes including B-series for entry level workloads, Dv3 and DSv3 series for mainstream production workloads, Esv3 series for memory intensive workloads and H-series for high performance computing needs. Azure supports up to 128 vCPUs and 504 GiB of RAM per VM.
Both platforms make it easy to scale compute capacity up or down to meet workload demands programmatically. Overall AWS offers a wider range of instance sizes for highly customizable sizing, while Azure provides simpler options that will meet most common business needs.
Storage
AWS offers simple object storage through S3, volume storage through EBS and archival storage through Glacier. Azure provides unstructured object storage through Blob storage, structured file storage through Azure Files, persistent volumes through Azure Disks and archive tier through Cool Blob storage.
A key difference is that Azure integrates object storage seamlessly with VM storage, while AWS treats them separately. Azure also offers higher durability SLAs on Blob storage compared to S3. But S3 features like object versioning and cross-region replication make it better suited for data governance needs.
Overall AWS storage services are more full-featured, cost effective and durable for cloud native workloads. But Azure provides the convenience of unified storage management across VMs and object storage.
Databases
Both AWS and Azure cater to a wide variety of database needs including relational/SQL, NoSQL, data warehouses, in-memory caches and time-series databases.
Key AWS database services are RDS (MySQL, PostgreSQL, SQL Server, Oracle), DynamoDB NoSQL database, ElastiCache in-memory caches, Redshift data warehouse and Timestream time-series database.
Azure offers Azure SQL Database relational databases, CosmosDB NoSQL database, Azure Cache for Redis in-memory caching solution, Synapse Analytics data warehouse, Azure Database for PostgreSQL and MySQL.
Third-party managed databases from vendors like MongoDB, Datastax and Redis are also available on both cloud platforms.
Overall both platforms have comparable capabilities for supporting database needs. AWS offers some specialized databases not available on Azure today like Graph, DocumentDB and QLDB. But Azure SQL and CosmosDB come with greater enterprise-grade capabilities out of the box.
Networking
AWS and Azure both provide comprehensive networking capabilities including hybrid connectivity, load balancing, DDoS protection, CDNs and traffic routing features.
Key AWS networking managed services are VPC for isolated cloud networks, Direct Connect and VPN for hybrid cloud access, Route 53 DNS routing, CloudFront content delivery and API Gateway.
Azure offers VNet virtual networking, ExpressRoute dedicated hybrid connectivity, Load Balancer for scale load distribution, Traffic Manager for global traffic routing, CDN for accelerated content distribution and API Management services.
A major Azure advantage for enterprises is the breadth of hybrid connectivity options including site-to-site VPN, private leased lines and ExpressRoute partnerships with >95% of telecom carriers for building low latency, secured connections.
Ecosystem
A huge reason behind AWS popularity has been their Marketplace featuring over 10,000 third party software offerings that can be easily deployed on AWS infrastructure. This includes databases, analytics tools, developer tools, management tools and just about every popular open source and commercial application.
Azure also provides thousands of certified software offerings through it‘s Marketplace. But the AWS ecosystem is comparatively much richer. AWS also has a strong partner network of system integration and technology partners that can support your cloud adoption journey.
So organizations that need the flexibility to leverage a variety of ISV and open source software should consider AWS as the platform of choice. Azure has been rapidly expanding application availability, but still has a lot of catching up to do.
Hybrid Cloud Capabilities
Hybrid cloud refers to an environment combining a private cloud platform (such as an on-premises data center) with a public cloud platform like AWS or Azure. Many enterprises choose hybrid cloud to maintain control over sensitive systems on-premise while taking advantage of cloud elasticity and economics for other workloads.
AWS capabilities for hybrid deployments has been one area where Azure has traditionally held an edge. Azure Stack and Azure Arc makes it easier for organizations to run Azure infrastructure on-premise and consistently manage services across cloud boundaries.
But Amazon has upped it‘s hybrid cloud game with AWS Outposts – rack-scale AWS hardware that can be deployed on-premise just like a server rack while seamlessly integrating with the AWS cloud environment. For true consistency across on-premise and cloud, Azure still holds the advantage. But AWS offers greater flexibility around on-premise hardware configurations that suit your needs.
Maturity & Reliability
As one of the oldest cloud computing platforms, AWS has great maturity in it‘s cloud infrastructure and services. It offers an industry leading uptime SLA of 99.99% for it‘s cloud infrastructure. Azure core infrastructure also provides a 99.99% monthly uptime SLA along with options for zone-redundant services and availability zones to make solutions resilient to data center failures.
Both platforms offer comprehensive security certifications including SOC 1/2/3, ISO 27001, PCI DSS level 1 and others. Azure has slightly higher default service limits in areas like number of cores, concurrent application requests etc that can support larger scale deployments with ease.
Overall, while AWS has greater production-use maturity, Azure offers well-proven enterprise grade reliability on par with AWS.
Pricing Comparison
AWS uses a pay-as-you-go pricing model without upfront fees or long term contracts. Users are billed on a per-second basis for use of compute, storage and managed services. Most services include a free usage tier. For eg: AWS offers 750 hours of Linux t2.micro instance usage per month for 12 months to all new accounts.
Azure also follows a pay-as-you-go model with per-second billing, monthly packages without contracts and no termination fees. Low priority VMs can be used at significant discounts. Free offers on Azure are more limited though – $200 worth of services for 30 days and $25 worth free for 12 months.
Azure does provide better cost transparency and analysis tools compared to AWS. The total cost of ownership can vary drastically based on instance types, data transfers, storage provisioning and complementary services used on top. But both platforms allow ways to optimize cloud spend including spot/low priority instances, auto-scaling, reserved capacity, storage lifecycle management and usage monitoring.
Hybrid cloud capabilities also reflect pricing differences. AWS Outposts adds the cost for on-prem hardware while Azure Stack is sold as an integrated on-premise solution. Azure Arc adds minimal costs to extend services on-premises.
For high demand workloads that allow maximizing spot instance or low priority VM usage, AWS tends to be significantly cheaper than Azure. But otherwise there is unlikely to be any major pricing benefit choosing one over the other.
Key Differences at a Glance
| Factor | AWS | Azure |
|---|---|---|
| Core design principles | Cloud native, programmable and flexible | Hybrid, enterprise-grade, integrated |
| Ease of getting started | Instant account setup, wide array of quickstart options | Some delays in account creation, focused quickstart guides |
| Main target audience | Tech startups, agile teams | Enterprise IT departments |
| Learning curve | Steeper due to more DIY options | Simpler ramp up for .NET apps |
| Hybrid capabilities | AWS Outpost racks allow hybrid deployments | Azure stack consistency across cloud boundaries |
| Network connectivity | Direct Connect using colocation partners | Rich connectivity integrations through ExpressRoute |
| Marketplace | 10,000+ software offerings | Curated, smaller selection of popular apps |
| Support experience | Documentation-driven. Concierge Support at a premium | Engineers assist across account types |
| Granular capabilities | More service options that can be combined in unique ways | Curated capabilities aligned to common architecture patterns |
Which Should You Choose?
We see that AWS and Azure share a lot of common capabilities around flexible compute, storage, networking, analytics and other services. They reflect different philosophies in their approach with AWS focusing on open programmability while Azure emphasizes an integrated enterprise platform.
AWS tends to be preferred by tech startups, companies with mainly Linux/open source software environments and teams seeking a high degree of customizability. The self-service capabilities, marketplace ecosystem and proven operational maturity makes it easy to get started on AWS.
Azure tends to be preferred by .NET centric organizations and companies that value smooth hybrid cloud capabilities. The tight integration between on-premise environments running Active Directory, Group Policy Manager and System Center and the Azure cloud platform is a key advantage here.
For most mid to large enterprises taking a cloud-first approach, evaluating workload needs and cost is crucial in the platform decision. Often a hybrid model across AWS and Azure services makes sense based on application architectures, compliance needs or geographic availability of certain capabilities. As capabilities evolve across platforms with unique differentiators, lock-in should also not be a concern during initial evaluations.
Summarizing Key Decision Factors
Consider AWS if…
- You value cloud native agility with a programmable environment
- Your workloads are mainly Linux or open source centric
- You need complex big data, ML and analytics capabilities
- Flexible and resilient scaling of infra and services is important
- You can maximize spot instance or low priority VM usage
- You want the richest breadth of technology integrations
Consider Azure if…
- You have significant Microsoft technology investments
- Compliance with regulatory controls is important
- You value smooth hybrid cloud capabilities
- Simplified IT resource management is key for your teams
- High availability redundancy with Availability Zones appeals to you
- Rich enterprise platform capabilities matter more than scale out
The Road Ahead
As the cloud industry evolves, AWS and Azure are positioned to leapfrog each other with new capabilities on a regular basis. Azure is focused on more industry vertical solutions leveraging IoT, edge compute and AI capabilities. AWS meanwhile works on reinventing cloud infrastructure itself for greater efficiency.
Multi-cloud adoption where organizations choose services from multiple providers is also on the rise. 82% of enterprises are expected to pursue a multi-cloud approach by 2022 [2]. This provides the freedom to solve for the unique technical and business needs of each application with agility.
In the foreseeable future, there might be workloads best-suited for Google Cloud Platform or AliBaba Cloud over AWS and Azure. Partnerships amongst cloud providers and platforms supporting multi-cloud deployments will help decide the right technology for your application need.
References
[1] Synergy Research Group
[2] LogicMonitor State of Multi-Cloud Report 2021