Major Cloud Platforms
AWS, Azure, GCP core services and global footprint comparison
AWS, Microsoft Azure, and Google Cloud Platform (GCP) are the three hyperscale cloud providers, collectively holding ~66% of public cloud market share (2024). AWS leads with the broadest service catalogue (~240+ services), Azure excels in enterprise Windows/.NET workloads and hybrid scenarios, while GCP differentiates on data analytics (BigQuery) and Kubernetes (GKE's origin). Understanding service equivalents across clouds is critical for multi-cloud governance and avoiding accidental lock-in.
Key Points
- AWS global infrastructure: 33 regions, 105 AZs (2024); Azure: 60+ regions; GCP: 40 regions.
- AWS object storage (S3) pioneered the industry; Azure Blob Storage and GCP Cloud Storage are functionally equivalent with minor API differences.
- Managed Kubernetes: EKS (AWS), AKS (Azure), GKE (GCP) — GKE is most feature-complete as Kubernetes originated at Google.
- Serverless compute: AWS Lambda (most mature), Azure Functions (deep .NET/Visual Studio integration), GCP Cloud Functions/Cloud Run.
- Data warehousing: AWS Redshift (columnar, RA3 nodes), Azure Synapse Analytics, GCP BigQuery (serverless, per-query billing) — BigQuery's serverless pricing model is a differentiator.
- IAM primitives differ: AWS uses IAM Roles + Policies (JSON), Azure uses Entra ID + RBAC (role assignments on scopes), GCP uses Service Accounts + IAM Bindings.
- Egress costs are a key financial lever — all three providers charge for data leaving the cloud region (~$0.09/GB), incentivizing single-cloud architectures.
- Marketplace offerings (AWS Marketplace, Azure Marketplace, GCP Marketplace) allow purchasing third-party software against cloud billing commitments.
| Service Category | AWS | Azure | GCP |
|---|---|---|---|
| Virtual Machines | EC2 | Virtual Machines | Compute Engine |
| Managed Kubernetes | EKS | AKS | GKE |
| Serverless Functions | Lambda | Azure Functions | Cloud Functions / Cloud Run |
| Object Storage | S3 | Blob Storage | Cloud Storage |
| Block Storage | EBS | Managed Disks | Persistent Disk |
| Managed SQL (Postgres) | RDS / Aurora | Azure Database for PG | Cloud SQL / AlloyDB |
| NoSQL (Key-Value) | DynamoDB | Cosmos DB | Firestore / Bigtable |
| Data Warehouse | Redshift | Synapse Analytics | BigQuery |
| Message Queue | SQS | Service Bus Queues | Pub/Sub |
| Event Streaming | Kinesis / MSK | Event Hubs | Pub/Sub Streaming |
| CDN | CloudFront | Azure CDN / Front Door | Cloud CDN |
| DNS | Route 53 | Azure DNS | Cloud DNS |
| IAM | IAM | Entra ID / RBAC | Cloud IAM |
| Secrets | Secrets Manager / KMS | Key Vault | Secret Manager / KMS |
| Observability | CloudWatch / X-Ray | Azure Monitor / AppInsights | Cloud Operations Suite |
| IaC Native | CloudFormation / CDK | Bicep / ARM | Deployment Manager / Config Connector |
Real-World Example
Airbnb migrated to a mostly-AWS architecture (EKS, RDS Aurora, S3, Kinesis) while retaining Snowflake (multi-cloud) for analytics — a common pattern of cloud-primary with best-of-breed SaaS.