AWS Insights
Powering the world’s largest events: How Amazon CloudFront delivers at scale
Behind every successful game launch, viral video, or global streaming event lies a question: How do you deliver content to millions of users simultaneously? Amazon CloudFront, AWS’s content delivery network (CDN), answers this challenge daily. This was most recently demonstrated by a record-breaking peak of 268 terabits per second (Tbps) on Nov. 1, 2025, driven by major game delivery workloads. To put this into context, that’s enough bandwidth to simultaneously stream live sports in HD to about 45 million concurrent viewers. This technical milestone demonstrates the scale and reliability that CloudFront and AWS bring to media, entertainment, and the gaming industry.
CloudFront Traffic Peak: 268 Tbps delivered seamlessly during a major game release, showcasing the scale and reliability of AWS’s content delivery network infrastructure.
With CloudFront, large-scale game downloads and video streaming delivery have become routine; what once required extensive planning and careful orchestration now happens seamlessly. For example, when a large gaming company releases updates for their most popular game title, or a top video streaming provider delivers popular sports to a record number of viewers, CloudFront handles the distribution to millions of concurrent users globally, allowing these companies’ development and content teams to focus on game creation and content production rather than delivery logistics.
CloudFront’s ability to handle such large-scale events has been made possible by continuous infrastructure innovation at AWS over the last 16 years, with each breakthrough building upon the last to create a platform that powers today’s Internet experiences worldwide.
The challenge of large-scale events
From the Super Bowl and Thursday Night Football in North America to the recent ICC Women’s Cricket World Cup reaching global audiences, millions of viewers expect a seamless streaming experience with no buffering, delays, or interruptions. Today’s digital events generate traffic volumes that would have been unimaginable just a few years ago, creating spikes that can overwhelm traditional content delivery systems. Even a few seconds of downtime during the Super Bowl or a major game launch can result in millions of frustrated users. CloudFront’s infrastructure and systems are designed to address this challenge.
CloudFront global edge network: Built for scale
CloudFront operates an extensive global edge network, infrastructure deployed at the outer boundaries of the AWS network where it connects to the Internet. This design specifically handles the massive traffic surges that come with large-scale events. However, scale isn’t just about having more servers; it’s about having the right infrastructure in the right places.
“With live sporting events, reliability and low latency are absolutely critical because every lost second negatively impacts viewers. Fans expect no buffering or interruptions, especially while watching live. Demand for live events is inherently dynamic—from scheduled kickoffs to unexpected game-defining moments—requiring infrastructure that scales instantly,” explains Ashish Ajwani, Head of Global Video Playback and Delivery, Amazon Prime Video. “Our AWS architecture leverages Amazon CloudFront’s elastic capacity and intelligent caching at multiple layers to seamlessly handle these spikes, allowing us to reliably deliver events like ‘Thursday Night Football’ and the ‘NBA on Prime’ to millions of viewers globally without interruption.”
Today, CloudFront operates 750+ edge locations, also known as points of presence (PoPs), across 440+ locations globally, delivering content with milliseconds of latency to viewers worldwide. These PoPs are strategic server locations at the edge of our network that cache and deliver content, helping ensure fast, reliable access for users anywhere in the world.
Through continuous innovation, we enhance our PoPs to increase processing power, network throughput, and cache storage. Our latest generation of PoPs delivers three times the performance of previous versions, enabling CloudFront to seamlessly handle millions of concurrent users, whether they’re downloading a highly anticipated game release or streaming a major sporting event.
Bringing content even closer with embedded PoPs
While CloudFront edge PoPs operate at the border of the AWS network and the Internet, some latency-sensitive workloads benefit from even closer proximity to end viewers for optimal performance. CloudFront embedded PoPs solve this problem. Embedded PoPs are deployed directly within Internet Service Provider (ISP) networks to cache large content catalogs closer to viewers, ensuring reliable delivery even during Internet congestion.
Today, CloudFront, in partnership with ISPs, operates 1140+ embedded PoPs across 100+ ISPs in 20+ countries. These PoPs are built specifically for delivering large-scale cacheable traffic such as live streaming, game downloads, and highly popular video-on-demand content. By placing content delivery infrastructure in the last mile of ISP networks, embedded PoPs reduce the distance that data needs to travel, improving performance for viewers, especially for those in areas further from major metropolitan centers.
ISPs of any size can request an embedded PoP through the CloudFront embedded PoP portal, or by contacting the AWS Peering and Interconnection team.
Protecting against threats while handling flash crowds
CloudFront’s intelligent systems can tell the difference between legitimate traffic surges and attacks. CloudFront leverages built-in protections and AWS Shield Standard to protect against known application layer offenders and network layer Distributed Denial of Service (DDoS) attacks at no additional cost. To further secure your workloads, you can enable additional services such as AWS Shield Advanced, AWS WAF, and AWS WAF Bot Control, which are co-hosted on the CloudFront edge locations.
CloudFront Flash Crowd Success Stories: Real-world examples of high-traffic events where CloudFront’s capacity management and intelligent routing delivered consistent performance during peak demand periods.
CloudFront’s capacity management system handles legitimate “flash crowds” – sudden, massive spikes in traffic that occur during events such as major game releases or live streaming events. The system constantly monitors global traffic patterns and infrastructure health, routing content through optimal locations based on capacity, proximity, and performance. This ensures that when millions of legitimate users request content simultaneously, they receive fast, reliable access even under extreme demand.
Planning your next global event
Whether you’re preparing for major sporting events, game launches, or other high-traffic events, CloudFront can help deliver your content with confidence and reliability. Contact your AWS account team to discuss event planning and capacity requirements. Our specialists will analyze your expected traffic, assess regional requirements, and ensure you are ready to handle peak demand while delivering a world-class viewing experience to your audience.