AWS Global Infrastructure¶

AWS operates one of the world's most resilient and widely distributed cloud networks. Understanding its physical layout enables you to design high-performance, fault-tolerant architectures.

graph TD
    Region["AWS Region (e.g., us-east-1, ap-south-1)"]
    AZ1["Availability Zone A (Data Center Cluster)"]
    AZ2["Availability Zone B (Data Center Cluster)"]
    AZ3["Availability Zone C (Data Center Cluster)"]
    Edge["Edge Locations / CloudFront PoPs (750+)"]

    Region --> AZ1
    Region --> AZ2
    Region --> AZ3
    Edge -.->|Caches Content from| Region

1. Regions¶

A Region is a separate geographic area (e.g., us-east-1 in N. Virginia, eu-west-1 in Ireland, ap-south-1 in Mumbai). Each region is completely autonomous and isolated from all other regions.

As of 2026, AWS operates 37+ Regions globally (including dedicated digital sovereignty regions such as the AWS European Sovereign Cloud, operated exclusively within the EU by EU-resident personnel).

How to Choose a Region¶

When launching resources, choose your Region based on:

Proximity to Users: Lower latency for your end users.
Data Sovereignty & Compliance: Legal requirements dictating that data must reside within specific national borders (e.g., GDPR in Europe, HIPAA in the US).
Service Availability: Newer services (e.g., specialized Bedrock models or cutting-edge EC2 instances) typically launch first in major regions (us-east-1, us-west-2, eu-west-1).
Cost: Pricing varies slightly by region based on local taxes and power costs.

2. Availability Zones (AZs)¶

Each AWS Region contains a minimum of three Availability Zones (e.g., us-east-1a, us-east-1b, us-east-1c).

An AZ consists of one or more physically distinct data centers equipped with independent power, cooling, and network links.
AZs within the same region are separated by a meaningful physical distance (up to ~60 miles / 100 km) to prevent correlated disaster impact (e.g., floods or local power grid outages), yet connected by ultra-low latency fiber to allow synchronous data replication.
AWS operates 100+ Availability Zones globally.

High Availability Rule of Thumb

Never deploy all your instances or database nodes in a single AZ. Distribute workloads across at least two AZs so that if an AZ suffers an unexpected outage, your application remains online.

3. Edge Locations & Points of Presence (PoPs)¶

AWS runs a global edge network of 750+ Points of Presence (including Edge Locations and Regional Edge Caches) in over 100 cities across 50+ countries.

Edge Locations do not run full virtual machines; they cache static and dynamic web content closer to users via Amazon CloudFront (CDN) and accelerate DNS lookups via Route 53.
This significantly reduces round-trip latency for global applications.

4. AWS Local Zones & Wavelength¶

AWS Local Zones: Extensions of AWS Regions placed in major metropolitan areas lacking a full Region (e.g., Chicago, Los Angeles, Buenos Aires), designed for sub-10ms latency applications such as real-time gaming, live video rendering, and financial trading.
AWS Wavelength: Embeds AWS compute and storage inside 5G telecommunications networks for ultra-low latency mobile edge computing.

Common Region Codes¶

Region Name	Identifier Code
US East (N. Virginia)	`us-east-1`
US West (Oregon)	`us-west-2`
Europe (Ireland)	`eu-west-1`
Europe (Frankfurt)	`eu-central-1`
Asia Pacific (Mumbai)	`ap-south-1`
Asia Pacific (Singapore)	`ap-southeast-1`
Asia Pacific (Tokyo)	`ap-northeast-1`
Asia Pacific (Sydney)	`ap-southeast-2`