The Portugal-Spain match at the 2026 World Cup revealed more than just sports news. It also set a new all-time high for traffic in the global interconnection ecosystem of DE-CIX: 28.4 terabits per second, recorded on 07/06/2026 at 19:35 UTC, 21:35 in Central European Time. This figure highlights how big live events have become a real test for networks, data centers, cloud platforms, CDNs, and interconnection providers.
DE-CIX compares this peak to over 15 million TikTok videos played simultaneously. The analogy is useful for understanding scale, but the key point is something else: traffic doesn’t grow smoothly during such an event. It appears almost suddenly. Millions of users connect within the same window from smart TVs, smartphones, tablets, and computers to watch exactly the same live content.
Stackscale (Aire), a European company specializing in cloud infrastructure, bare-metal, and private cloud, has also analyzed this phenomenon on its blog. The takeaway is relevant for any digital business: the World Cup streaming isn’t just a matter for operators and major audiovisual platforms, but a reminder that infrastructure must be designed for peaks, not just comfortable average loads.
Live events change traffic patterns
On-demand consumption spreads out the load more evenly. A series episode, a software update, or a backup can be distributed over hours or days. A World Cup match, however, doesn’t work that way. Demand concentrates just before kickoff, during key plays, at halftime, and in the final minutes.
This simultaneity completely alters the required architecture. It’s not enough to have capacity for average traffic. You need to be able to absorb very intense bursts without degrading the experience. In video, that means avoiding freezes, slow starts, sudden quality drops, or excessive delays compared to live broadcast.
DE-CIX also points out that growth isn’t only driven by the main video stream. Around the match, there is AI-generated summaries, real-time stats, social media, second-screen experiences, personalized content, and interactive platforms for fans. All of these add traffic and demand low latency, high availability, and quick responsiveness.
For a company, comparing that to the World Cup might seem exaggerated, but the pattern is the same. An e-commerce site on Black Friday, ticket sales, product launches, viral campaigns, auctions, exam-season educational platforms, or SaaS with concentrated usage windows can face similar issues but on a smaller scale.
Neutral points, CDNs, and data centers: the unseen infrastructure
When a user clicks “watch live,” they expect the video to work. Behind that is a complex chain: access providers, backbone networks, CDNs, cloud platforms, data centers, streaming operators, and Internet exchange points.
Internet exchange points, or IXPs, enable different networks to exchange traffic directly. This reduces unnecessary routing, improves latency, and increases efficiency. DE-CIX emphasizes that these points are crucial for connecting Internet providers, cloud providers, CDNs, and streaming platforms during peak demand periods.
Stackscale echoes this idea, explaining that Internet performance depends not only on “the cloud” as an abstract concept but also on physical infrastructure like racks, routers, switches, fiber optics, interconnection ports, storage, and distributed computing platforms.
In a global event, bringing content closer to users is vital. The less data has to travel, the better the experience. That’s why CDNs distribute content across multiple locations, and well-connected data centers play a central role. Computing power helps, but connectivity often determines whether the service can sustain the load.
Spain also part of the infrastructure
DE-CIX notes that the company operates three neutral points in the Iberian Peninsula: Madrid, Lisbon, and Barcelona. Along with Marseille and Palermo, the DE-CIX Southern Europe ecosystem connects over 500 networks across 17 data centers. Established in 2016, DE-CIX Madrid has become a digital hub within that network, managing traffic flows of more than 1.5 Tbit/s.
This is important because Internet infrastructure isn’t only organized around major global tech capitals. Connected regions, local hubs, nearby data centers, and neutral networks directly affect the end-user experience. For Spanish and European businesses, having close, well-interconnected infrastructure can make a difference in latency, control, data sovereignty, and predictability.
Discussions about infrastructure tend to arise when something goes wrong — a service crashes, a website slows down, a stream cuts out, or a campaign can’t handle demand. But preparation should happen beforehand. The architecture is truly tested when many users do the same at the same time.
What businesses should learn
The record of 28.4 Tbit/s doesn’t mean every company must scale like a global streaming platform. But it does offer practical lessons.
First, averages can be misleading. A service may have a reasonable monthly consumption yet fail during a critical hour. Designing solely based on averages can hide bottlenecks in databases, storage, load balancers, firewalls, caches, or the network.
Second, adding servers isn’t always the solution. Sometimes the problem is a slow query, saturated storage, a heavy frontend, a misconfigured CDN, or an external dependency that doesn’t scale at the same pace. Stackscale summarizes this well: before indiscriminately adding capacity, review architecture, caches, databases, media, physical limits, and saturation points.
Third, dedicated infrastructure remains relevant for critical loads. Public cloud offers elasticity, but not all workloads benefit equally. For applications with intensive consumption, sustained performance needs, low latency, or predictable costs, options like private cloud, bare-metal, or hybrid architectures can provide better control.
Fourth, high availability isn’t improvised. Simply replicating virtual machines isn’t enough. You need to define RPO, RTO, disaster recovery plans, backups, periodic testing, monitoring, alerts, and service dependencies.
The World Cup exemplifies the most demanding Internet scenario: millions of users connected simultaneously, high-quality video, synchronized consumption, and additional services around the broadcast. For companies, the core idea is simple: infrastructure shouldn’t only be built to perform well when everything is normal but also to keep functioning reliably when demand peaks.
Frequently Asked Questions
What does reaching 28.4 Tbit/s in traffic mean?
It means DE-CIX’s global interconnection ecosystem recorded an aggregated peak of 28.4 terabits per second during the Portugal-Spain match at the World Cup.
Why does sports streaming cause such intense peaks?
Because millions of users connect almost simultaneously to watch the same live content, unlike on-demand video, which distributes load more evenly.
What role do Internet exchange points play?
They enable networks, operators, CDNs, cloud providers, and digital platforms to exchange traffic directly, reducing unnecessary routes and improving latency.
What can a company learn from this record?
It should design its infrastructure for actual peaks, not just average traffic. This involves reviewing network, compute, storage, caches, databases, load balancing, and observability.
Why mention Stackscale in this context?
Because Stackscale analyzes the impact of massive streaming from an infrastructure point of view, focusing on cloud, data centers, interconnection, latency, and readiness for traffic peaks.
