Building resilience: Inside AWS’s nine million kilometers of fiber-optic cabling

A hidden network of fiber-optic cables keeps the world connected. When you send an email across continents or stream a video from another country, that data travels through physical infrastructure underground and under the sea. These cables—many no bigger than a garden hose—are part of the physical foundation of the internet that billions depend on daily. Amazon Web Services (AWS) has built over nine million kilometers of network infrastructure, so vast it could reach from Earth to the Moon and back more than 11 times.

Businesses, governments, and billions of people now depend on continuous connectivity for vital systems, financial networks, healthcare delivery, and daily communications, making the resilience of this network more essential than ever. AWS Global Infrastructure spans 38 AWS Regions and 120 Availability Zones, all connected through multiple redundant pathways with built-in resiliency designed to meet this demand for reliable connectivity.

We take a measured approach to building and expanding this infrastructure. Our network expansion follows careful analysis of customer needs and usage patterns. This ensures we build the right capacity in the right places at the right time, while preparing for our customers’ future needs. As more customers move towards multi-region deployments for enhanced scale, availability, and global reach, we continue to strategically evolve our network. We deliver multiple layers of redundancy and fault tolerance to provide increased resilience.

Under the sea

The AWS network operates across multiple diverse routes in every major subsea corridor worldwide. By operating various routes, we support customer access to their cloud workloads and data, regardless of where they operate globally. For the past decade, through direct investments and strategic partnerships, we’ve systematically expanded our subsea backbone. We regularly review our network topology and diversity, strategically adding new paths. This topology review involves analyzing our network’s physical layout and connection patterns to identify potential vulnerabilities and ensure we maintain multiple independent pathways between regions—a critical capability that allows us to spot single points of failure and strategically invest in additional routes to deliver high availability.

At AWS, we “design with failure in mind.” This principle ensures customers can depend on our network for their mission-critical applications and data access. By proactively designing for multiple unpredictable failures, we help customers maintain business continuity even during network disruptions. What makes AWS’s approach distinctive is our robust infrastructure design: We maintain diverse network pathways and alternative connectivity options, while our systems are carefully designed to isolate failures and prevent cascading impacts across regions.

We monitor and assess potential risks to our subsea cables continuously after installation. Risk mitigation involves education and awareness programs with stakeholders and seabed users, such as through collaboration with organizations like the European Subsea Cable Association (ESCA) and the International Cable Protection Committee (ICPC), and through direct engagement with local fishing communities where our cables operate. We also advocate for increased industry investment in marine repair capabilities as part of our comprehensive approach to network resilience. In areas where these measures may not be sufficient, or where we face unavoidable natural and human risks—such as seismic activity in tectonically active regions or high-traffic shipping lanes where anchor strikes are more likely—AWS invests in additional cable routes to support consistent service. Through our extensive redundancy, if one pathway is compromised, our network automatically reroutes traffic to help maintain connection stability.

A clear demonstration of this resilience came in 2024, when four subsea cables were damaged off the coast of Abidjan in the Ivory Coast due to a landslide. This coincided with separate cable system damage in the Red Sea. Through our network diversity strategy, which includes carefully selected alternative routes up the west side of Africa, AWS maintained service even when primary paths were compromised. During this period of widespread disruption, our customers in Cape Town continued to access their cloud resources.

Under the ground

While our subsea infrastructure continues to expand, it works hand-in-hand with our ever-growing terrestrial network. Each new data center brings a new point of connection to the broader AWS network, regularly expanding our infrastructure by thousands of miles of fiber to bring cloud resources closer to where customers operate. AWS continues to rapidly scale its data center capacity worldwide to meet the growing demand for AI compute power. What sets AWS apart is our integrated approach to terrestrial infrastructure. AWS designs a cohesive global network where every new connection strengthens the entire system’s resilience.

To keep up with customer demand for high-performance computing and data-intensive AI workloads, from model training to large-scale inference, we optimize our fiber infrastructure through strategic hardware selection and upgrades. Much of this work happens at the optical interconnect level. The vast majority of our network capacity now uses 400 Gigabit Ethernet (GbE) technologies—four times faster than the previous standard of 100 GbE. AWS invests heavily in both the hardware and software that runs our optical transceivers, which are crucial for high-speed data transfer. These devices convert electrical signals to light for transmission over fiber optic cables and vice versa, connecting devices within our data centers and establishing high-bandwidth, low-latency connections to our customers’ networks. We continue to upgrade our entire network to faster and more reliable optics. Such innovations extend to the fiber cables themselves, with technologies like hollow-core fiber optics offering 30% improvement in transmission latency, and multi-core fiber increasing data transmission capacity.

Throughout our global infrastructure, including our network backbone, AWS integrates robust security measures, from physical protection to end-to-end cybersecurity protocols. Our network incorporates multiple layers of security controls to protect customer data as it moves across our infrastructure. We apply encryption at the network level to all traffic that passes outside of AWS control between regions, adding an additional layer of protection for customer data across our global network. AWS provides customers with the tools and capabilities to implement encryption-in-transit. We offer TLS encryption, VPN connections, and dedicated network connections through AWS Direct Connect so customers can protect their most sensitive workloads—from financial transactions to healthcare records to proprietary AI models. Customer data encrypted before it enters our network will maintain that encryption throughout its journey, whether traveling through subsea cables across oceans or terrestrial fiber across continents. This gives customers control over their encryption and confidence that their data remains protected during transmission. Physical safeguards include strategic cable burial and restricted facility access, while our security systems provide protection for the applications and workloads running on the network. For example, AWS leverages threat intelligence from a massive graph model with billions of nodes to detect and thwart an average of 124,000 new malicious internet domains daily. Similarly, AWS uses data from its unique global network of sensors to analyze over 100 million interactions a day and automatically mitigate threats. And AWS threat intelligence tools identify and shut down over 7 billion daily unauthorized external scanning attempts of AWS services within minutes, at global scale.

The increasing popularity of AI and ML workloads continues to create greater demand for AWS services, adding significantly more traffic across our network. We analyze vast amounts of network usage data to proactively deploy additional capacity where customers need it most. As global digital infrastructure becomes increasingly important, we’re actively investing in strategic ways to expand our network backbone, supporting businesses as they build and scale their most demanding applications. AWS takes a long-term view of global connectivity requirements. We’re not just building for today’s connectivity demands—we’re architecting the foundation for tomorrow’s digital innovations, from quantum computing to immersive technologies that haven’t been invented yet. Looking ahead, we recognize the importance of streamlining the process for building new capacity to meet future needs. We’re proactively engaging with policymakers and industry partners worldwide to develop efficient frameworks for deploying this important infrastructure. By collaborating on innovative solutions, we aim to enhance global connectivity and support technological advancement, helping build a robust digital future that benefits everyone.