Editorial note, updated October 2026: Edgegap's median server deployment time is 2 seconds, from cold start. That means provisioned from scratch rather than allocated onto an already-running server, and measured to container ready, before the game engine starts (rolling 30 days, as of 18 September 2026). The numbers below are from 2021 and are kept as a historical record. Live figure and methodology: platform data page

As you may know, we’re working on pretty cool stuff here at Edgegap. Today we’d like to share some data on scaling and deployment times using our different solutions.

Launching a new online multiplayer game is a daunting task. You need to make sure the experience within these first few hours of online services can. But at the same time, you need to be able to scale your game hosting infrastructure so that you don’t leave new players on the sidelines.

Here at Edgegap, we use a mix of edge computing and container technology to deploy game servers on demand, where it is needed, when it is needed. This is very different from the current status-quo of having a base pool of game servers rented on a yearly basis, and complementing that with an “on-demand” cloud offering. I purposefully put on-demand between commas here because it is not a truly on-demand experience; you will need to wait several minutes for the cloud provider to give you access to the instance you requested, and will also need to configure it correctly before you can use it.

In February 2021, on-demand at Edgegap meant an average of 6.7 seconds before your game server was running where you needed it. That number came from data we gathered over a single day with 500,000 deployments.

Optimizing for speed is what we do at Edgegap, and being able to launch game servers as fast as possible where the players are has been a point of focus here for the past year. Every single step of the process had to be optimized to save every millisecond possible to make sure that the players don’t have to wait.

Talking about speed, Edgegap was founded on the promise of reducing latency for players. At the time, we could choose from a pool of 220 locations, with the average latency for these 500,000 deployments hovering at around 35ms. This is only possible by bringing game servers where the players are, and deploying these game servers on-the-fly with our container-based solutions.

So if you’ve been looking for a better way to host game servers to better serve your players, look no further. Get in touch to talk about your projects, and we’ll be there to accompany you on the journey!