Most hosting companies rent their network reach from the same upstream providers their customers are already trying to reduce dependency on. InMotion Hosting took a different approach, and a recent infrastructure upgrade makes that distinction more concrete than it has been before.
The company completed a significant overhaul of its global network by integrating a software-defined networking fabric into its existing peering architecture across three locations: Los Angeles, Ashburn in Virginia, and Amsterdam. The upgrade transforms each site from a fixed hardware-bound peering hub into a programmable networking gateway that connects directly to cloud providers, carriers, internet exchanges, and managed service partners without requiring new physical hardware for every new relationship.
The practical difference is in how connections multiply. Traditional internet exchange peering runs one physical port per relationship. Adding a new cloud provider or carrier means buying and provisioning another port, which takes time and money. The SDN fabric replaces that model with a single high-capacity port that splits into multiple virtual cross connects, each pointing to a different destination, with bandwidth adjustable in software rather than rebuilt in hardware.
On top of that fabric sits InMotion’s existing direct peering with Google, Microsoft, and Cloudflare across its three locations. Together, those relationships cover the majority of traffic flowing to US and European audiences. Traffic moving across those private paths stays off the public internet entirely, which removes variable hop counts and reduces exposure in a single architectural decision.
For customers running latency-sensitive workloads, the upgrade shows up in the places that are easy to feel. Checkout flows move faster. API calls return more consistently. Database round-trips follow shorter, more predictable routes. Dedicated server customers get up to 10 Gbps port speeds with actual headroom rather than overage fees when demand spikes.
Erik Soroka, InMotion’s Director of IT and Data Center Operations, noted that the company’s engineers can now observe real-time global metrics and respond directly when latency increases or route instability appears, rather than waiting on an upstream provider to investigate and act.
InMotion has owned and operated its own infrastructure since 2001. That history matters here because the upgrade builds on network assets the company controls rather than capacity leased from someone else.
