AWS History and Timeline regarding AWS Global Infrastructure - Regions, Availability Zones, Summary of Expansions, and Introduction
First Published:
Last Updated:
While AWS publishes the current list of Regions and Availability Zones, the historical launch dates are spread across dozens of individual announcements. This article consolidates them into a single chronological reference: the general availability date of every AWS Region from the first Region in 2006 through the newest Regions and the AWS European Sovereign Cloud, together with the structural milestones — Availability Zones, AWS Local Zones, AWS Wavelength, and AWS Outposts — that define how AWS infrastructure is organized.
This article focuses on Region-level launches and major infrastructure milestones, not on the enumeration of every edge location or on which services are available in which Region. For per-service history, see the companion timeline series; for the current, authoritative counts and per-Region service availability, always refer to the official AWS Global Infrastructure page.
This is part of the series that began with the "AWS History and Timeline - Almost All AWS Services List, Announcements, General Availability(GA)", in which I extract features from the history and timeline of AWS. Related individual timelines include Amazon EC2, Amazon VPC, and Amazon CloudFront.
I am summarizing the present-day structure of the infrastructure as a Current Overview, Structure, and Components of AWS Global Infrastructure.
The figure below shows, at a glance, how AWS Global Infrastructure spread geographically by era — from a single Region in 2006 to a worldwide footprint across six continents.
Background and Method of Creating the AWS Global Infrastructure Historical Timeline
The purpose of this timeline is:- Tracking the history of AWS Global Infrastructure and organizing the sequence of Region launches and structural milestones.
- Summarizing the definitions and characteristics of the components (Regions, Availability Zones, Local Zones, Wavelength Zones, edge locations, and Outposts) so that the dates can be read consistently.
The main sources referenced for this timeline are the official AWS announcements:
- What's New with AWS?
- AWS News Blog
- AWS Regions and Availability Zones documentation and its document history
- AWS Global Infrastructure (for the current, authoritative overview)
Definitions used in this timeline (the consistency of these definitions is the core of this reference):
- Region: A physical location in the world that contains a cluster of data centers, organized into multiple, isolated Availability Zones. Each AWS Region is a separate geographic area with independent infrastructure and its own API endpoint identifier (for example,
ap-northeast-1). - Availability Zone (AZ): One or more discrete data centers within a Region, each with redundant power, networking, and connectivity, housed in separate facilities. AZs in a Region are physically separated by a meaningful distance yet close enough for low-latency synchronous replication. An AZ is identified by its Region code plus a letter (for example,
ap-northeast-1a), and each AZ also has an account-independent AZ ID (for example,apne1-az1) that refers to the same physical location across accounts. - Local Zone: An extension of an AWS Region that places compute, storage, and a subset of services in a large metropolitan or industry center to deliver single-digit-millisecond latency to nearby users.
- Wavelength Zone: AWS compute and storage embedded inside a communications service provider's 5G network, at the edge of the mobile network, for ultra-low-latency mobile and edge applications.
- Edge location / Point of Presence: Sites used by Amazon CloudFront, AWS Global Accelerator, and Amazon Route 53 to cache content and terminate connections close to end users. The growth of the edge network is summarized only briefly here; its detail is covered in the Amazon CloudFront timeline.
- AWS Outposts: Fully managed racks and servers that bring AWS infrastructure and services to on-premises and edge locations.
- Dedicated Local Zones: Local Zones dedicated to a specific customer or community, built to meet data isolation, sovereignty, and compliance requirements.
- AWS European Sovereign Cloud: A physically and logically separate cloud for Europe, operated entirely within the European Union with its own partition and Regions.
Method: For each Region, the date recorded here is the date AWS announced general availability — the day the Region became usable by customers — as stated in the official AWS announcement. Earlier "in the works," "coming soon," or limited-preview announcements are mentioned in context where relevant, but the launch date is the general-availability date. Where an official announcement states the number of Availability Zones at launch, that count is included. Some announcement pages are dated in one time zone while regional press may cite the adjacent calendar day; in those cases the date from the primary AWS announcement is used.
Please note that the items on this timeline are not every change to AWS Global Infrastructure, but representative milestones — each Region's general availability plus the introduction of major infrastructure constructs — that I have selected for the overview and the definitions above.
A note on the AWS China Regions: The AWS China Regions operate differently from all other AWS Regions. To comply with China's laws and regulations, they are operated by licensed local partners — the Beijing Region (
cn-north-1) by Beijing Sinnet Technology Co., Ltd. (Sinnet), and the Ningxia Region (cn-northwest-1) by Ningxia Western Cloud Data Technology Co., Ltd. (NWCD). They use a separate set of account credentials and an isolated partition, and are not accessible with a standard global AWS account.How to read the dates in this timeline: A few situations deserve explanation so that each date is interpreted correctly.
- Preview versus general availability: Some Regions were announced in a limited preview before they were generally available. The clearest example is the China (Beijing) Region, announced in preview in December 2013 and generally available in 2016. In such cases this timeline places the row at the confirmed announcement and describes the general-availability step in the text, rather than inventing a precise general-availability day that is not documented in a primary AWS source.
- Local Region versus full Region: The Asia Pacific (Osaka) location appears twice — once as a single-Availability-Zone Local Region in 2018, and again in 2021 when it was expanded into a standard, three-Availability-Zone Region. Both are shown because the change of form is itself a meaningful milestone.
- Partitions and account systems: Most Regions belong to the standard (commercial) AWS partition and are reachable with a normal AWS account. The China Regions, AWS GovCloud (US) Regions, and the AWS European Sovereign Cloud sit in separate partitions with their own account systems, which is why they are called out explicitly.
- Availability Zone counts: Where an official announcement states the number of Availability Zones at launch, that count is recorded. Older announcements frequently did not state a number; modern Regions typically open with three Availability Zones. A Region's Availability Zone count can also grow after launch, and those later additions are not tracked row-by-row here.
AWS Global Infrastructure Historical Timeline (Region Launches from August 25, 2006)
The following table lists, in ascending order, the general availability of each AWS Region together with the major structural milestones of AWS Global Infrastructure. Each row links to the official AWS source. Use the year index to jump to a period.2006 | 2008 | 2009 | 2010 | 2011 | 2012 | 2013 | 2014 | 2016 | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 | 2025 | 2026
| Date | Summary |
|---|---|
| 2006-08-25 | The first AWS Region, US East (N. Virginia), is established (us-east-1). Amazon EC2 launched as a limited public beta in a single US East location that AWS later formalized as the US East (N. Virginia) Region. At this point the terms "Region" and "Availability Zone" were not yet in use; they were introduced as the model matured. [Source] [Doc history] |
| 2008-03-26 | Availability Zones and Elastic IP addresses are introduced for Amazon EC2. This established the Region-and-Availability-Zone model — multiple isolated Availability Zones within a Region — that underpins AWS fault isolation and high availability to this day. [Source] |
| 2008-12-10 | Europe (Ireland) opens as the first AWS Region outside the United States (eu-west-1), with 2 Availability Zones. Announced as "Amazon EC2 Crosses the Atlantic." Amazon S3 had begun offering storage in Europe about a year earlier (2007), but this was the first full compute Region outside the US. [Source] |
| 2009-12-03 | US West (N. California) Region opens (us-west-1). AWS's third Region, giving customers a West Coast option in the United States. [Source] |
| 2010-04-29 | Asia Pacific (Singapore) opens as AWS's first Region in Asia Pacific (ap-southeast-1). AWS's first infrastructure presence in Asia. [Source] |
| 2011-03-02 | Asia Pacific (Tokyo) Region opens (ap-northeast-1). AWS's fifth infrastructure location worldwide and second in Asia Pacific, giving Japanese customers in-country low-latency capacity. [Source] |
| 2011-08-16 | AWS GovCloud (US) launches (us-gov-west-1). The first AWS Region isolated for US government workloads with specific regulatory requirements such as ITAR. It was later renamed GovCloud (US-West) when a second GovCloud Region followed. [Source] |
| 2011-11-09 | US West (Oregon) Region opens (us-west-2). A second US West Coast Region, priced to match US East (N. Virginia). [Source] |
| 2011-12-14 | South America (São Paulo) opens as AWS's first Region in South America (sa-east-1). [Source] |
| 2012-11-12 | Asia Pacific (Sydney) Region opens (ap-southeast-2). AWS's third Region in Asia Pacific and its first in Australia. [Source] |
| 2013-12-18 | AWS announces the China (Beijing) Region in limited preview (cn-north-1). This became AWS's first infrastructure in mainland China. It reached general availability in 2016, operated by Beijing Sinnet Technology Co., Ltd. (Sinnet) to comply with China's regulations, using a separate account system and partition. [Source] [China Gateway] |
| 2014-10-23 | Europe (Frankfurt) Region opens (eu-central-1). AWS's second Region in Europe, adding an option in continental Europe with German data residency. [Source] |
| 2016-01-06 | Asia Pacific (Seoul) Region opens with 2 Availability Zones (ap-northeast-2). [Source] |
| 2016-06-27 | Asia Pacific (Mumbai) opens as AWS's first Region in India, with 2 Availability Zones (ap-south-1). [Source] |
| 2016-10-17 | US East (Ohio) Region opens with 3 Availability Zones (us-east-2). With this launch AWS's footprint reached 14 Regions and 38 Availability Zones, and it gave customers a second, lower-latency option in the eastern United States. [Source] |
| 2016-12-08 | Canada (Central) opens as AWS's first Region in Canada (ca-central-1). [Source] |
| 2016-12-13 | Europe (London) Region opens (eu-west-2). AWS's third Region in Europe, with UK data residency. [Source] |
| 2017-12-11 | AWS China (Ningxia) Region opens (cn-northwest-1), operated by NWCD, with 2 Availability Zones. AWS's 17th Region globally and its second in China, operated by Ningxia Western Cloud Data Technology Co., Ltd. (NWCD), structurally parallel to Sinnet operating the Beijing Region. [Source] |
| 2017-12-18 | Europe (Paris) Region opens (eu-west-3). AWS's 18th Region and fourth in Europe. [Source] |
| 2018-02-13 | The Asia Pacific (Osaka) Local Region opens (ap-northeast-3) — AWS's first-ever Local Region, with a single Availability Zone. An isolated, invite-only facility intended to be paired with the Tokyo Region for in-country disaster recovery. It was later expanded into a full Region in 2021. [Source] |
| 2018-11-12 | AWS GovCloud (US-East) Region opens with 3 Availability Zones (us-gov-east-1). AWS's second GovCloud Region, giving US government workloads an East Coast option. [Source] |
| 2018-12-12 | Europe (Stockholm) Region opens with 3 Availability Zones (eu-north-1). AWS's fifth Region in Europe and its first in the Nordics. [Source] |
| 2019-04-25 | Asia Pacific (Hong Kong) Region opens with 3 Availability Zones (ap-east-1). [Source] |
| 2019-07-30 | Middle East (Bahrain) opens as AWS's first Region in the Middle East, with 3 Availability Zones (me-south-1). [Source] |
| 2019-12-03 | AWS Local Zones launch, beginning with Los Angeles. Local Zones place compute and storage in a major metropolitan area to deliver single-digit-millisecond latency to nearby users, as an extension of a parent Region. [Source] |
| 2019-12-03 | AWS Outposts becomes generally available. Fully managed racks that bring AWS infrastructure and services on-premises, extending the Region to a customer's own data center. Outposts had been announced a year earlier at AWS re:Invent 2018. [Source] |
| 2020-04-22 | Africa (Cape Town) opens as AWS's first Region in Africa, with 3 Availability Zones (af-south-1). [Source] |
| 2020-04-27 | Europe (Milan) Region opens with 3 Availability Zones (eu-south-1). AWS's sixth Region in Europe. [Source] |
| 2020-08-06 | AWS Wavelength becomes generally available. The first Wavelength Zones open in Boston and the San Francisco Bay Area on Verizon's 5G network, embedding AWS compute at the edge of the mobile network for ultra-low-latency applications. Wavelength had been announced at AWS re:Invent 2019. [Source] |
| 2021-03-01 | The Asia Pacific (Osaka) Region is expanded from a Local Region into a full, standard Region with 3 Availability Zones (ap-northeast-3). The number of Availability Zones increased from one to three and the service portfolio expanded, making Osaka Japan's second full Region. [Source] |
| 2021-12-13 | Asia Pacific (Jakarta) opens as AWS's first Region in Indonesia, with 3 Availability Zones (ap-southeast-3). [Source] |
| 2022-08-29 | Middle East (UAE) Region opens with 3 Availability Zones (me-central-1). AWS's second Region in the Middle East. [Source] |
| 2022-11-08 | Europe (Zurich) Region opens with 3 Availability Zones (eu-central-2). AWS's 28th Region, adding Swiss data residency. [Source] |
| 2022-11-15 | Europe (Spain) Region opens with 3 Availability Zones (eu-south-2). [Source] |
| 2022-11-21 | Asia Pacific (Hyderabad) opens as AWS's 30th Region and second in India, with 3 Availability Zones (ap-south-2). [Source] |
| 2023-01-23 | Asia Pacific (Melbourne) Region opens with 3 Availability Zones (ap-southeast-4). AWS's second Region in Australia. [Source] |
| 2023-08-01 | Israel (Tel Aviv) Region opens with 3 Availability Zones (il-central-1). [Source] |
| 2023-12-20 | Canada West (Calgary) opens as AWS's 33rd Region, with 3 Availability Zones (ca-west-1). AWS's second Region in Canada. [Source] |
| 2024-08-21 | Asia Pacific (Malaysia) Region opens with 3 Availability Zones (ap-southeast-5). [Source] |
| 2025-01-07 | Asia Pacific (Thailand) opens as AWS's first Region in Thailand, with 3 Availability Zones (ap-southeast-7). [Source] |
| 2025-01-14 | Mexico (Central) opens as AWS's first Region in Mexico, with 3 Availability Zones (mx-central-1). [Source] |
| 2025-06-05 | Asia Pacific (Taipei) opens as AWS's first Region in Taiwan, with 3 Availability Zones (ap-east-2). AWS stated that this expanded its footprint to 117 Availability Zones across 37 geographic Regions. [Source] |
| 2025-09-01 | Asia Pacific (New Zealand) opens as AWS's first Region in New Zealand, with 3 Availability Zones (ap-southeast-6). [Source] |
| 2026-01-14 | The AWS European Sovereign Cloud opens, with its first Region in Brandenburg, Germany (eusc-de-east-1). A physically and logically separate cloud for Europe, operated within the EU with multiple Availability Zones and its own partition. It was announced in 2023 with an original target of the end of 2025. [Source] |
Note: In parallel with Region growth, AWS also expanded its edge network (Amazon CloudFront edge locations and Points of Presence), AWS Local Zones (which expanded internationally starting in 2022), and AWS Dedicated Local Zones for sovereignty workloads. Those are summarized here only at the milestone level; the edge network's detailed history is covered in the Amazon CloudFront timeline.
Key Themes in the Evolution of AWS Global Infrastructure
The table above lists each launch as a discrete event. Reading the timeline as a whole, several themes emerge that explain how AWS Global Infrastructure reached its present shape. Each theme below is a synthesis of the dated milestones already listed, not a new set of claims.The founding era and the birth of the Region and Availability Zone model (2006 to 2012)
AWS began in 2006 with a single location on the US East Coast, and at first there was no formal notion of a "Region" or an "Availability Zone" at all — the earliest Amazon EC2 beta simply ran in one place. The concepts that now define the platform were introduced in March 2008, when Availability Zones (together with Elastic IP addresses) gave customers a way to run across multiple, physically isolated data centers inside one location. That single design decision — grouping several isolated Availability Zones into one Region — became the foundation for everything that followed, because it let AWS offer high availability within a Region while keeping Regions independent from one another.With the model in place, AWS expanded outward. Europe (Ireland) in December 2008 was the first Region outside the United States; US West (N. California) followed in 2009; and Asia Pacific (Singapore) in 2010 gave AWS its first presence in Asia. The year 2011 was pivotal, adding Asia Pacific (Tokyo), the first isolated GovCloud Region for US government workloads, a second US West Region in Oregon, and South America (São Paulo). By the time Asia Pacific (Sydney) opened in 2012, AWS had a Region on five continents, all built on the same Region-and-Availability-Zone pattern established in 2008.
Global expansion accelerates and diversifies (2013 to 2017)
The next phase broadened both the geography and the operating models. The China (Beijing) Region, announced in preview in 2013 and generally available in 2016, introduced a distinctive arrangement in which the Region is operated by a licensed local partner to satisfy national regulations. Europe (Frankfurt) in 2014 added a second European Region with strong data-residency appeal, and 2016 brought a burst of growth — Seoul, Mumbai (the first Region in India), Ohio (a second Region in the eastern US, which took the footprint to 14 Regions), Canada (Central), and London. In 2017, the China (Ningxia) Region and Europe (Paris) rounded out an 18-Region footprint. This era is where data residency and in-country presence became explicit drivers of where AWS built, rather than simply following demand for raw capacity.Reaching new continents and closing gaps (2018 to 2021)
The following years filled in parts of the map AWS had not yet reached. The Asia Pacific (Osaka) Local Region in 2018 was a first-of-its-kind construct — a single-Availability-Zone facility meant to be paired with Tokyo for in-country disaster recovery — and a second GovCloud Region (US-East) and Europe (Stockholm) also arrived that year. In 2019, Asia Pacific (Hong Kong) and, significantly, Middle East (Bahrain) — AWS's first Region in the Middle East — opened. In 2020, Africa (Cape Town) became AWS's first Region on the African continent, alongside Europe (Milan). By 2021, AWS had expanded Osaka from a Local Region into a full three-Availability-Zone Region, and opened Asia Pacific (Jakarta) in Indonesia. The theme of this era is completeness: covering continents and countries that had previously relied on Regions elsewhere.The edge and hybrid era: Local Zones, Wavelength, and Outposts
Alongside full Regions, AWS built constructs that push capacity beyond the traditional Region boundary. AWS Local Zones, launched in Los Angeles in December 2019 and later expanded to many metropolitan areas, place a subset of services close to users for single-digit-millisecond latency while remaining tied to a parent Region. AWS Wavelength, announced at the same period and generally available in August 2020, embeds compute inside carriers' 5G networks for ultra-low-latency mobile and edge applications. AWS Outposts, generally available in December 2019, brings AWS-managed racks and servers into a customer's own data center. Together these mean "where AWS runs" is no longer only a list of Regions; it also includes metropolitan Local Zones, carrier-network Wavelength Zones, on-premises Outposts, and the global edge network used by services such as Amazon CloudFront.The sovereignty era: partner-operated Regions, GovCloud, and the European Sovereign Cloud
A recurring thread across the whole timeline is sovereignty and regulatory isolation. It appears early with AWS GovCloud (US) in 2011 and the partner-operated China Regions, and it becomes a defining theme in the most recent years. AWS Dedicated Local Zones offer Local Zones reserved for a single customer or community with additional isolation. The AWS European Sovereign Cloud, announced in 2023 and opened with its first Region in Brandenburg, Germany, in January 2026, is the fullest expression of this idea: a physically and logically separate cloud with its own partition, operated entirely within the European Union. For architects, the practical consequence is that "which Region" is increasingly a question about legal jurisdiction and operational control, not only latency.How the pace and shape of expansion changed
Read end to end, the timeline shows the cadence of expansion increasing over time. The first five Regions took from 2006 to 2011; by the mid-2020s, AWS was opening multiple Regions per year — for example, Thailand, Mexico, Taipei, and New Zealand all opened between January and September 2025. The mix also shifted: early Regions frequently launched with two Availability Zones, whereas modern Regions typically open with three, reflecting a standardized design for resilience. And the definition of a "location" broadened from a Region to a spectrum that now spans full Regions, Local Zones, Wavelength Zones, Outposts, the edge network, and dedicated sovereign clouds.Current Overview, Structure, and Components of AWS Global Infrastructure
AWS Global Infrastructure is organized as a hierarchy of Regions and Availability Zones, extended by edge and hybrid constructs that place capacity closer to users or on-premises. The sections below summarize how the pieces fit together and the considerations for choosing where to run workloads.AWS Global Infrastructure Use Cases
- Selecting the primary Region for a new application based on proximity to users, data-residency requirements, and service availability.
- Designing multi-Availability-Zone architectures that survive the failure of a single data center within a Region.
- Building multi-Region architectures for disaster recovery, business continuity, or global active-active operation.
- Meeting data-residency, sovereignty, and regulatory requirements by keeping data within a specific country or jurisdiction (including the AWS China Regions and the AWS European Sovereign Cloud).
- Reducing latency for end users with AWS Local Zones, and for mobile and 5G edge applications with AWS Wavelength.
- Extending AWS to on-premises and edge sites with AWS Outposts for low latency, local data processing, or data residency.
Specific Examples of Use Cases
- A company subject to Japanese data-residency requirements runs its workloads in the Asia Pacific (Tokyo) Region and uses the Asia Pacific (Osaka) Region for in-country disaster recovery.
- A European public-sector organization with the strictest sovereignty needs deploys to the AWS European Sovereign Cloud.
- A media company distributes streaming to global viewers through Amazon CloudFront edge locations while origin content is served from a small number of Regions.
- A game studio places latency-sensitive session servers in an AWS Local Zone near a major city, backed by the parent Region.
- A manufacturer runs low-latency control systems on AWS Outposts inside a factory while managing them through a nearby Region.
Key Components of AWS Global Infrastructure
- Region: An isolated geographic area containing multiple Availability Zones, with an independent set of infrastructure and its own API endpoint identifier. Regions are isolated from one another by default so that a problem in one Region does not affect another.
- Availability Zone (AZ): One or more discrete data centers within a Region, each with redundant power, networking, and connectivity. Deploying across multiple AZs is the primary way to build highly available applications within a Region.
- AZ ID: A consistent, account-independent identifier for a physical Availability Zone (for example,
use1-az1), useful when coordinating zone placement across accounts. - AWS Local Zones: Extensions of a Region placed in additional metropolitan areas for single-digit-millisecond latency.
- AWS Wavelength: AWS infrastructure embedded in 5G carrier networks for ultra-low-latency mobile and edge use cases.
- AWS Outposts: Fully managed AWS racks and servers operated in a customer's own data center or edge location.
- Edge locations and Points of Presence: The global network used by Amazon CloudFront, AWS Global Accelerator, and Amazon Route 53 to serve content and route traffic close to users.
- AWS Dedicated Local Zones: Local Zones dedicated to a specific customer or community for data isolation and compliance.
- AWS GovCloud (US): Isolated Regions for US government and regulated workloads, accessed with separate credentials.
- AWS European Sovereign Cloud: A separate cloud and partition for Europe, operated entirely within the EU.
How Regions, Availability Zones, and the Global Edge Fit Together
These components form a hierarchy. At the top are Regions, which are isolated from one another by design: by default, a Region operates independently, and services and data are scoped to a Region unless you explicitly use a cross-Region feature. Within each Region are multiple Availability Zones, which are the unit of fault isolation you design around for high availability — a well-architected application typically runs across at least two or three Availability Zones so that the loss of a single data center does not take the application down.Extending outward from the Region are the constructs that trade some of a Region's completeness for proximity. Local Zones and Wavelength Zones are attached to a parent Region but sit physically closer to users or inside a carrier's network, offering a subset of services with lower latency. AWS Outposts brings AWS hardware into your own facility while still being managed through a Region. Finally, the global edge network — the edge locations and Points of Presence used by Amazon CloudFront, AWS Global Accelerator, and Amazon Route 53 — caches content and terminates connections close to end users worldwide, in front of whatever Regions host the origin.
Understanding this hierarchy is the key to using the timeline in practice: choosing a Region sets your baseline for jurisdiction, latency, and service availability; spreading across Availability Zones gives you resilience within that Region; adding a second Region gives you disaster recovery or global reach; and Local Zones, Wavelength, Outposts, and the edge network let you meet latency or on-premises requirements that a Region alone cannot.
AWS Global Infrastructure Region Selection Considerations
When choosing where to deploy, architects typically weigh:- Latency: Proximity of the Region (or a Local Zone or Wavelength Zone) to end users.
- Data residency and sovereignty: Legal or policy requirements to keep data within a country or jurisdiction, including the China Regions and the AWS European Sovereign Cloud.
- Service and feature availability: Not every service or feature is available in every Region at launch; the current per-Region availability should be checked on the official AWS pages.
- Resilience: The number of Availability Zones in a Region and the option to add a second Region for disaster recovery.
- Cost: Prices and data-transfer charges vary by Region; consult the official AWS pricing pages (pricing is intentionally not reproduced here).
As of this article's Last Updated date (2026-07-05), AWS spans dozens of Regions across six continents, and it continues to grow. As specific, source-anchored reference points from the timeline above: the Asia Pacific (Taipei) launch in June 2025 brought AWS to 117 Availability Zones across 37 geographic Regions, the Asia Pacific (New Zealand) Region followed in September 2025, and the AWS European Sovereign Cloud opened its first Region on a separate partition in January 2026. Because the total is a moving target, this article does not assert a single current count; for the authoritative, up-to-date number of Regions and Availability Zones, see the official AWS Global Infrastructure page.
Frequently Asked Questions about AWS Global Infrastructure History
When did AWS launch its first Region?
AWS's first Region, US East (N. Virginia) (us-east-1), traces to August 25, 2006, when Amazon EC2 launched as a limited public beta in a single US East location. The formal Region-and-Availability-Zone model came slightly later: Availability Zones were introduced for Amazon EC2 on March 26, 2008.
When did the AWS Tokyo Region open?
The Asia Pacific (Tokyo) Region (ap-northeast-1) opened on March 2, 2011. It was AWS's fifth infrastructure location worldwide and its second in Asia Pacific, after Singapore.
What is the difference between a Region and an Availability Zone?
A Region is an isolated geographic area (for example, Europe (Ireland) or Asia Pacific (Tokyo)) that contains multiple Availability Zones. An Availability Zone is one or more discrete data centers within that Region, each with redundant power, networking, and connectivity, physically separated from the other zones yet close enough for low-latency replication. You choose a Region for latency and data residency, and you spread workloads across Availability Zones within the Region for high availability.
When did AWS launch Local Zones and Wavelength?
AWS Local Zones launched on December 3, 2019, beginning with Los Angeles, and later expanded to many more metropolitan areas. AWS Wavelength was announced at AWS re:Invent 2019 and became generally available on August 6, 2020, with the first Wavelength Zones in Boston and the San Francisco Bay Area on Verizon's 5G network.
How many AWS Regions are there?
The number grows over time, so a single fixed figure quickly becomes outdated. As source-anchored reference points: AWS stated 117 Availability Zones across 37 geographic Regions at the Asia Pacific (Taipei) launch in June 2025; the Asia Pacific (New Zealand) Region followed in September 2025; and the AWS European Sovereign Cloud opened a first Region on a separate partition in January 2026. For the current, authoritative count, refer to the official AWS Global Infrastructure page.
Why are the AWS China Regions operated differently?
To comply with China's laws and regulations, the AWS China Regions are operated by licensed local partners: the Beijing Region (cn-north-1) by Beijing Sinnet Technology Co., Ltd. (Sinnet), and the Ningxia Region (cn-northwest-1) by Ningxia Western Cloud Data Technology Co., Ltd. (NWCD). They use a separate account system and an isolated partition, and are not accessed with a standard global AWS account.
Which was the first AWS Region on each continent?
AWS's first Region was US East (N. Virginia) in North America, in 2006. The first Region outside the United States was Europe (Ireland), in 2008. The first in Asia Pacific was Asia Pacific (Singapore), in 2010. The first in South America was South America (São Paulo), in 2011. The first in the Middle East was Middle East (Bahrain), in 2019. The first in Africa was Africa (Cape Town), in 2020.
What is the difference between a Local Zone, a Wavelength Zone, and an AWS Outpost?
All three place AWS capacity outside a standard Region, but for different purposes. A Local Zone is an AWS-operated extension of a Region in a major metropolitan area, for low-latency access to nearby users. A Wavelength Zone embeds AWS compute inside a communications provider's 5G network, for ultra-low-latency mobile and edge applications. An AWS Outpost is a fully managed rack or server installed in your own data center or facility. Local Zones and Wavelength Zones run in AWS or carrier facilities, whereas Outposts run on premises under AWS management.
Summary
From a single location in 2006, AWS Global Infrastructure has grown into a worldwide network of Regions and Availability Zones spanning six continents, extended by Local Zones, Wavelength Zones, Outposts, an edge network, and a dedicated European Sovereign Cloud. This timeline captures the general availability date of each AWS Region — from US East (N. Virginia) and the first Region outside the US in Ireland, through Tokyo, the China Regions, the first Regions in Africa and the Middle East, the expansion of Osaka from a Local Region to a full Region, and the newest Regions in Malaysia, Thailand, Mexico, Taipei, and New Zealand — alongside the structural milestones that shaped how AWS organizes capacity.For per-service history, see the overview and related timelines: the AWS History and Timeline overview, Amazon EC2, Amazon VPC, and Amazon CloudFront. To put this infrastructure to work across Regions, see the AWS multi-Region active-active architecture guide and the AWS disaster recovery strategies guide.
This timeline will be updated as AWS Global Infrastructure continues to evolve. For the current, authoritative list of Regions and Availability Zones, always refer to the official AWS Global Infrastructure page.
References:
AWS Global Infrastructure
AWS Regions and Availability Zones (Documentation)
AWS Availability Zones (Documentation)
Document history for AWS Regions and Availability Zones
What's New with AWS?
AWS News Blog
AWS China Gateway
Opening the AWS European Sovereign Cloud (AWS News Blog)
AWS History and Timeline - Almost All AWS Services List, Announcements, General Availability(GA)
AWS History and Timeline regarding Amazon EC2
AWS History and Timeline regarding Amazon VPC
AWS History and Timeline regarding Amazon CloudFront
AWS Multi-Region Active-Active Architecture Guide
AWS Disaster Recovery Strategies Guide
References:
Tech Blog with curated related content
Written by Hidekazu Konishi