Looking glass
Run traceroutes and pings from any of our 67 locations. See exactly how your traffic routes from our network.
The network global platforms run on.
- 85 Tbps
daily traffic peak
- 3,000+
ISPs reached globally
- 67
global locations
- 50
backbone-connected DCs
Network architecture
330+ Tbps. Built to reach the networks your users are on.
Three components: our backbone carries traffic between regions, peering hands it off into your user’s network, and transit covers whatever we do not peer with.
- 15 Tier‑1 transit providers
Orange, Arelion, Lumen, Liberty Global, GTT, and 10 more. BGP optimization selects the best path based on performance, not cost.
- 300+ directly peered networks
A private interconnect means one hop from us into the network your user is on. Fewer networks in the path, fewer places to congest at peak.
- 50 Tbps private backbone
A server in Paris serving a user in New York crosses the Atlantic on our own circuits, not handed off to transit at the nearest exchange. 400GE and 100GE circuits across the 50 data centers the backbone connects.
The last hop
Worldwide scale. Local precision.
300+ networks peered directly, chosen location by location, with over 3,000 ISPs reachable through those interconnects and our transit. The last hop, from our network into the network your user is on, is where peak congestion shows up. Everything before it runs on our own backbone.
Peered where your users are
Each data center carries the PNIs that matter in its region, so traffic takes the local path, not a detour through a distant hub. A server in Tokyo reaches SoftBank, NTT, and KDDI directly. Frankfurt reaches Deutsche Telekom, Vodafone, and O2. Twelve years of agreements, region by region.
Same standard in every location
Capacity sized to local demand with headroom on top. No location running on a single transit link. Same hardware standard, same SLA, same engineers in São Paulo as in Frankfurt.
Strategic peering at major internet exchanges
Alongside our direct PNIs, we peer at major regional internet exchanges including AMS-IX, DE-CIX, LINX, and HKIX. That expands local reach and gives us more routing options into the networks your users depend on.
Peered where your users are
Each data center carries the PNIs that matter in its region, so traffic takes the local path, not a detour through a distant hub. A server in Tokyo reaches SoftBank, NTT, and KDDI directly. Frankfurt reaches Deutsche Telekom, Vodafone, and O2. Twelve years of agreements, region by region.
Same standard in every location
Capacity sized to local demand with headroom on top. No location running on a single transit link. Same hardware standard, same SLA, same engineers in São Paulo as in Frankfurt.
Strategic peering at major internet exchanges
Alongside our direct PNIs, we peer at major regional internet exchanges including AMS-IX, DE-CIX, LINX, and HKIX. That expands local reach and gives us more routing options into the networks your users depend on.
Cloud connectivity
Direct connections to AWS, GCP, and Azure.
Many workloads run across dedicated infrastructure and cloud at the same time. We peer directly into the major clouds, so traffic between your servers and your cloud environment takes a short path.
- Google Cloud
Direct PNI in 23 locations: Amsterdam, Frankfurt, London, New York, Paris, Singapore, Sydney, Tokyo, and more.
- AWS
Direct PNI in 9 locations, including Ashburn, Frankfurt, London, New York, Paris, Prague, Singapore, and Vienna.
- Microsoft Azure
Direct PNI in 5 locations across Europe, North America, and Asia-Pacific.
Reliability and security
Built for reliability
Instead of pretending that hardware doesn’t fail and outages don’t exist, we come well prepared.
Availability
Network and power uptime are each committed at 99.99% monthly. Redundant routing and switching in every location, with multiple transit providers per data center.
DDoS protection
Shield is enabled by default in every location, not sold separately for dedicated servers. Mitigation in under 10 seconds for L4 attacks, volume attacks, and TCP SYN floods.
Learn moreMonitoring
Continuous last‑mile monitoring with automatic rerouting around congestion. Real‑time data, daily routing adjustments.
Network tools
See it for yourself.
Our looking glass and latency graphs are public, so every number above is measurable from outside.
Latency graphs
See how we connect to major networks, per location. Real measured latency, not theoretical specs.
Public status page
Real incident history and per-location uptime. The same data our team watches.
The network behind 67 locations, and every user beyond them.
15 transit providers300+ PNIs50 Tbps private backbone