[PLEU-NANO]
20 GbpsDedicated
- Processor
- 1 × E5-2620 v3
- Memory
- 64 GB DDR4
- Storage
- 50 GB SSD
- Est. setup
- 1–2 days
€1,499 / mo
OrderConcurrent traffic and headroom
20 Gbps Servers
Keep overlapping broadcasts on one server when their combined traffic no longer fits a 10 Gbps port. Compare the hardware below against the events that run at the same time.
Workload illustration, not live server measurements.
Server configurations
8 packages EUR monthly
The port is the server connection speed. It does not specify a minimum sustained rate, a monthly transfer allowance or an overage price. Ask for these terms for the exact package before ordering.
20 GbpsDedicated
€1,499 / mo
Order20 GbpsDedicated
€1,549 / mo
Order20 GbpsDedicated
€1,599 / mo
Order20 GbpsDedicated
€1,649 / mo
Order20 GbpsDedicated
€1,749 / mo
Order15 GbpsDedicated
€1,299 / mo
Order20 GbpsDedicated
€1,699 / mo
Order20 GbpsDedicated
DDoS protection: 2 TB total across the mitigation network. This is a volume, not a per-server mitigation rate. Confirm limits for this IP before ordering.
Check traffic terms for this package€1,699 / mo
OrderA 20 Gbps port is the next step when 10 Gbps would sit too close to saturation during the busiest hour. It is useful for live events, growing IPTV relays, and regional cache nodes with uneven demand.
Confirm the physical interface, single-flow limits and routing before ordering. A 20 Gbps package name alone does not tell you how traffic is distributed across interfaces.
Use cases
Two simultaneous 6 Gbps events need 12 Gbps before overhead. The same events at different times need only 6 Gbps at peak. Use the actual schedule when sizing bandwidth for live events.
Workload: One node serving a city, country, or ISP-heavy audience from cached content.
Sizing: Cache hit ratio and storage size decide whether traffic is mostly outbound or also origin-heavy.
Fit: 20 Gbps gives meaningful headroom without moving into a large 40 Gbps footprint.
Limit: If cache misses are frequent, origin capacity matters as much as edge capacity.
Workload: Predictable bursts around matches, broadcasts, launches, or scheduled streams.
Sizing: Concurrency rises quickly; stream bitrates should be modeled before buying.
Fit: The extra headroom above 10 Gbps reduces peak-hour pressure.
Limit: For unpredictable spikes, pooling or 40 Gbps may be a better fit.
Workload: One server handling media delivery, API/static assets, and occasional transcoding.
Sizing: Separate disks for hot media and system work help keep the node responsive.
Fit: 20 Gbps suits mixed workloads where the port must not be the first bottleneck.
Limit: CPU-heavy transcode farms should be sized from compute first.
Provisioning
For several concurrent services, check container runtime support and network driver support before copying an existing system image.
Check installation, driver and licensing requirementsQuestions & answers
Use the port shown on the individual package. The 20 Gbps range also lists a 15 Gbps option; its lower port ceiling still applies when you choose it from this page.
Not necessarily. The client connection, round-trip time, packet loss, storage and TCP behaviour can limit one transfer. Test both a single download and several concurrent downloads from a representative client network.
Add simultaneous peaks, not daily totals. Two audiences drawing 8 Gbps and 6 Gbps at the same time need 14 Gbps before overhead. If they peak at different hours, measure the combined peak instead.
The speed alone does not specify the interface or bonding arrangement. Ask for the NIC, link layout and any per-flow limit before designing around a particular topology.
It describes the mitigation network, not guaranteed clean traffic for one server. Ask which traffic types are covered and what happens to your IP during an incident.
Related paths
Compare adjacent bandwidth, media, and pooled-capacity routes before choosing a configuration.