Blog
Blueprint: Brazil Hub, Regional PoPs, Smarter Routing
South America is no longer a “far edge.” Mobile internet users in Latin America almost doubled from about 230 million to nearly 400 million between 2014 and 2021. Platforms looking at the region may already run their main origins in Europe or the U.S. — and now need a CDN footprint in South America that doesn’t collapse under cross-continent latency.
For those teams, the strategy hinges on three levers: PoP placement across South America, inter‑country routing around a Brazilian hub, and origin consolidation with Europe/US as global cores and Brazil as the regional shield. The rest is plumbing.
South America CDN— Pre-reserve São Paulo: coming soon — CDN PoPs across 5 LATAM countries — 21 Tier III & IV DCs, incl. planned |
Which South American PoP Layout Best Reduces Cross-Continent Latency?
Evaluate São Paulo as a planned regional origin or customer-managed shield for a South America CDN with existing origins in Madrid or Atlanta. Melbicom’s current South American edge PoPs are Fortaleza, Rio de Janeiro, Lima, Bogotá, and Buenos Aires. Choose paths using measurements from the access networks your audience uses.
IX.br’s trajectory explains the role of Brazil. In March 2026, the national IXP fabric hit 50 Tbit/s of aggregate traffic, including 32 Tbit/s at IX.br São Paulo. These figures underscore São Paulo’s role as an interconnection hub; actual application performance still depends on the paths to your audience.
CDN planning in South America must factor in undersea routes. EllaLink reports round-trip delay below 60 ms between Sines and Fortaleza over its direct Brazil–Portugal connection, and below 120 ms between Madrid and São Paulo. These point-to-point figures help evaluate long-haul routes; end-user RTT and CDN TTFB require separate measurements.
For a pan‑regional CDN rollout in South America, a pragmatic pattern looks like this:
- Customer-managed shield in São Paulo – to consolidate cache misses from regional PoPs and pull from Madrid or Atlanta origins once capacity becomes available.
- Regional PoPs in Fortaleza, Rio de Janeiro, Lima, Bogotá, and Buenos Aires – existing Melbicom PoPs in these metros can shorten last‑mile paths and reduce dependence on long‑haul routes.
- Optionally, cable‑adjacent PoPs near landing points (for example on the Brazilian coast) when your cross‑continent flows are especially latency‑sensitive.
The aim is to serve users from regional PoPs and reduce unnecessary long-haul traffic; validate the actual edge and origin paths from the networks your audience uses.
How Should Traffic Be Routed Between Countries Around Brazilian Origin Hubs?
South American CDN routing should favor regional paths where they improve measured performance. The provider selects edge PoPs; cache misses can pass through a customer-managed São Paulo shield before reaching Madrid or Atlanta origins. Treat that shield as a planned architecture, and verify paths from each target country’s access networks.

The failure mode you want to avoid is classic: a user in Argentina hits content hosted in Spain or the U.S. via a convoluted path that bounces through Miami or random Tier‑2s, adding avoidable long-haul delay to content requests. Properly run, a CDN service in South America does the opposite:
- Geographic request routing at the edge to direct users toward suitable PoPs, with delivery paths verified for each target market.
- Brazil as a regional origin hub in a planned architecture: regional PoPs can fetch from a customer-managed shield in São Paulo, which in turn talks to Madrid and Atlanta origin clusters over tested long-haul paths.
- Local and regional peering across Latin American IXPs to keep inter‑PoP and PoP‑to‑Brazil traffic inside the continent wherever possible.
Origin routing control matters. For your origins, BGP policies can influence route selection through supported communities and available routes. Melbicom’s BGP Session service provides free BGP sessions on dedicated servers, IPv4/IPv6, BYOIP, and support for full, default, or partial routes at operating data center locations. These controls apply to your server network; CDN edge selection and internal BGP remain provider-managed.
Brazil Deploy Guide— Avoid costly mistakes — Real RTT & backbone insights — Architecture playbook for LATAM |
![]() |
Resilience rides on multi‑origin thinking. Practically, that means:
- Maintaining independent origin clusters (for example, Madrid + Atlanta) when the application’s availability requirements call for them.
- Configuring CDN origin pools and health checks, then testing failover when an origin cluster becomes unavailable.
- Testing alternative paths for BGP-enabled origins, including regional and long-haul failure scenarios.
From a South American user’s perspective, the priority is consistent, low-variance response time—whether content comes from a cache in Lima or an origin in Madrid is an implementation detail.
Which KPIs and Traffic Forecasts Should Drive a South America CDN Strategy?
The KPIs that actually move the needle in South America are latency, cache hit ratio, throughput headroom, and availability, all tied to realistic cross‑continent traffic forecasts. Design for video‑heavy, mobile‑first demand, then add capacity and PoPs based on measured TTFB, cache offload, and how quickly long‑haul links saturate.
Latency. Measure p95 RTT and TTFB separately by country, access network, and cache status. Set targets from application requirements and measured paths. Point-to-point cable figures from EllaLink help compare long-haul routes; application TTFB also includes connection and server-response time.
Cache and origin offload. For a modern CDN in South America, set cache-hit targets by content type and cacheability, and distinguish request hit ratio from byte offload. Dynamic or personalized content may correctly bypass caching. Investigate unexpected changes in TTLs, cache keys, and content reuse. The practical metric: what share of traffic is served from regional PoPs or a shield, and how much still reaches Madrid or Atlanta?
Capacity and per‑server bandwidth. That scale needs serious ports. In Madrid, Melbicom’s dedicated servers support ports up to 40 Gbps per server, with 90+ ready-to-go configurations. Our Atlanta facility supports up to 200 Gbps per server with 21 stock builds, ideal as a U.S. Southeast origin anchor. Overall, Melbicom’s footprint covers 21 Tier III/IV data centers, including planned São Paulo capacity, alongside CDN PoPs in 40 locations.
Availability and control. Finally, measure per‑PoP uptime, time to fail over during incidents, and the routing controls available to you. BGP sessions with BYOIP at operating dedicated-server locations support customer-origin routing, while CDN caching, origin pooling, and geographic request routing are configured separately.
Key Metrics and Considerations for a South America CDN Strategy
| Metric/Factor | Target / Example | Strategic Impact |
| Latency (RTT and TTFB) | p95 RTT and TTFB by market, access network, and cache status | Validate regional and long-haul paths; network RTT and application TTFB measure different parts of the delivery experience. |
| Cache Hit Ratio | Workload-specific request hit ratio and byte offload | Size origin capacity around cacheable content, deliberate bypasses, and measured long-haul traffic. |
| Peak Throughput | Load-tested forecast peak plus failover headroom | Size origin ports and CDN delivery capacity from event demand, cache-miss load, and failure scenarios. |
| Traffic Mix & Growth | Measured video share, request growth, and event peaks | Use workload-specific forecasts to size caching, origin capacity, and long-haul bandwidth. |
| Availability & Control | Application availability objective and tested failover time | Use CDN origin pooling and delivery controls; apply BGP/BYOIP to customer-operated origins. |
Key Design Takeaways for a Pan-Regional CDN South America Footprint
Once you map both infrastructure and demand, a few design patterns show up again and again for CDN South America deployments.
- Localize Aggressively and Peer in-Region – Evaluate regional peering at IX.br and national exchanges so South America CDN flows can avoid unnecessary detours through distant transit hubs. Confirm the effect with route and latency measurements.
- Use Brazil as the Regional Hub, Not the Only Origin – Keep Madrid and Atlanta as your global origin anchors while evaluating São Paulo as a regional customer-managed shield for South America. The shield consolidates cache misses from regional PoPs, reducing repeated long-haul origin fetches without a full origin migration.
- Design Inter-Country Routing Deliberately – Model flows between neighbors, not just “user to origin.” On your origins, use supported BGP communities to influence available paths, and measure forward and return paths for latency-sensitive apps.
- Size for Video and Peaks, Not Averages – For video-heavy workloads, plan capacity around peak events—launch days, finals, flash sales—and validate origin ports and CDN delivery capacity against forecasts and failover tests.
- Let KPIs Drive Expansion – Track TTFB per country, cache hit ratio per PoP, and link utilization into Brazil. Expand CDN coverage with the provider, or plan workloads for forthcoming São Paulo dedicated capacity, when you see those metrics drift, not when users start complaining.
Operationalizing a South America CDN Strategy: Brazil Hub + PoPs

Putting this into practice usually follows a three‑step arc:
- Phase 1 – Core Deployment – Stand up your initial origin cluster on dedicated servers in an available location such as Madrid or Atlanta, front it with the CDN, and configure origin pooling and Origin Shield for the workload. Plan São Paulo as a future regional origin or customer-managed shield when capacity becomes available. Establish BGP sessions plus BYOIP where your own origin-routing requirements call for them. Start with South America CDN PoPs that match your initial go‑to‑market countries.
- Phase 2 – Measurement and Policy Tuning – Instrument p95 RTT and TTFB, cache-hit ratio, and per-country concurrency. Review CDN delivery paths with the provider and tune origin BGP policies where applicable. Adjust cache TTLs and cache rules around traffic patterns (e.g., match‑day surges, trading windows, prime‑time streams).
- Phase 3 – Expansion and Hardening – As traffic grows, expand CDN coverage with the provider and add origin capacity in high-growth markets. Introduce secondary origins where the risk profile justifies them, and test multi-origin failover.
Validate the resulting performance and availability from each target market as traffic and routing change.
Deploying and Scaling with Melbicom
We at Melbicom built our platform for exactly this playbook: global origins plus configurable CDN delivery.

Melbicom’s single-tenant dedicated-server footprint spans 21 Tier III/IV data centers across Africa, Asia, Europe, North America, and South America, including planned São Paulo capacity. Our operating locations are backed by a 14+ Tbps backbone, 23 transit providers, and 29 IXPs. At operating locations, we keep 1100+ ready-to-go configurations in stock and deliver custom servers in 3–5 business days. Globally, Melbicom offers up to 200 Gbps per server. Melbicom’s CDN spans 40 PoPs in 35 countries, including Fortaleza, Rio de Janeiro, Lima, Bogotá, and Buenos Aires.
If you’re planning or refining your CDN service in South America and want to combine European/African/American origins with a South American edge offering configurable caching, origin pools, and delivery policies, contact us. We’ll help you translate traffic forecasts and KPI targets into concrete server reservations in existing data center locations, plus planning and pre-reservations for upcoming São Paulo capacity.
Deploy Your South America CDN
Build a regional edge in Brazil with origins in Europe or the U.S. Our team can map current CDN coverage, origin routing, and planned São Paulo capacity to your traffic targets across LatAm.
