ls /global-gateway
Low-latency, high-capacity global interconnect.
Private connectivity across Latitude.sh regions — no setup required.
./why-global-gateway
Always-on Private Backbone
Global Gateway is the fastest way to move workloads across Latitude.sh regions. With up to 400Gbps of intercontinental, private and dedicated bandwidth, pre-configured links between our data centers, your traffic never has to touch the internet.
Latency You Can Count On
Our backbone connects the most in-demand regions with enterprise-grade latency and is quickly expanding to reach all sites.
Latency | Ashburn | NYC | Dallas | Frankfurt | Tokyo | Singapore | Miami |
|---|---|---|---|---|---|---|---|
| Ashburn | — | 6 ms | 36 ms | 85 ms | — | — | — |
| NYC | 6 ms | — | — | 81 ms | 168 ms | — | — |
| Dallas | 36 ms | — | — | — | — | — | 34 ms |
| Frankfurt | 85 ms | 81 ms | — | — | 152 ms | — | — |
| Tokyo | — | 168 ms | — | 152 ms | — | 75 ms | — |
| Singapore | — | — | — | — | 75 ms | — | — |
| Miami | — | — | 34 ms | — | — | — | — |
Optimized network for distributed workloads
Low latency
Optimized regional routes.
High capacity
Ready for heavy workloads.
Always-on
Works out of the box.
Privacy and security
No internet exposure.
Dedicated Gateway
Reserved capacity when you need it.
No public IPs required
Keep your workloads inside a private RFC1918 segment. With Global Gateway, you can create cross-connects between regions and run workloads entirely on a private RFC1918 segment. That means no need for public IPs or exposure to external carriers.
./how-it-works
Shared by default, Dedicated on demand
All Latitude.sh locations are interconnected through a Shared Global Gateway, automatically enabled for every customer. No setup, no configuration — your servers can communicate across regions out of the box.
Need isolation or guaranteed throughput? Request a Dedicated Global Gatewayto connect your workloads across sites with predictable performance. A dedicated interface ensures your private traffic never competes with the public one — so latency-sensitive applications stay consistent.