Two edge computing case studies: Happier gaming and faster site development

In a recently published case study, edge software startup Edgegap is boasting that it dramatically improved the online game experience for customers of an unnamed game studio.

Case studies are carefully crafted marketing pieces and need to be digested accordingly. Deciding not to publish the studio’s name is further reason to weigh what is presented.

In this case, the Canadian startup is offering some positive performance numbers (not necessarily surprising for Edgegap, but the online game infrastructure industry).

Typically, gamers play across networks involving distant cloud installations, accruing the latency that comes with longer distances. Edgegap’s software, Arbitrium, instead chooses the closest edge site for each game instance. There being more edge sites, distances usually are shorter. Edgegap was able to select between 142 locations on which to run its software; the studio used six “relay” sites to enhance game performance.

Arbitrium reportedly improved latency for 91% of players using the unnamed studio games.

The software was able to cut the average round-trip time (the time it takes to send a packet to a destination and get it back) per player by 46.5% compared to using the cloud. Similarly, Edgegap said it reduced average round trip time per match by 36%.

A quarter of players reportedly saw latency fall below 50 milliseconds with Edgegap, compared with just 2% of players experiencing that level of performance through cloud infrastructure. Sixty-nine percent of players on Edgegap infrastructure saw latency below 100 milliseconds, compared to one-third experiencing that performance by playing through the cloud, according to Edgegap.

For reference, 100 milliseconds is about how long it takes to blink an eye – and the approximate time it takes for the brain to process an image. Enough time, in other words, for a game to be won or lost.

Edgegap also applied a fairness metric with which to judge its software’s performance. Not surprisingly, it measures the difference in latency between players. One player with higher lag than others is going to be unhappy, and likely to talk about it a lot on game forums.

In this case study, fairness was improved 66% of the time by cutting round-trip time. The difference in latency was cut by 31 milliseconds on average, the company claimed.

Hemnet develops real estate site with Cloudflare Workers

In a second new case study, the owners of Hemnet, a Swedish classified-ad site for those interested in buying, renting or brokering real estate, turned to Cloudflare for help opening up a new site section for people looking at property outside Sweden.

Unlike the utilitarian main site, the new section – Utland – would frame Hemnet’s typical search services with a more narrative approach. According to Cloudflare, Swedes wanted to learn about areas and about how property markets work before tucking in for a determined search.

Cloudflare was already Hemnet’s security and system performance vendor, so site owners chose its Workers solution. Workers is a tidy JavaScript execution environment through which customers make, deploy and scale serverless applications on Cloudflare’s 200 city network. Applications run on the closest data center.

According to Cloudflare, a small product team was able to launch Utland from the main domain in an amount of time that Hemnet leaders felt was quick. As a bonus, Hemnet has had more than 100,000 cyberthreats rebuffed each month by Cloudflare’s Advanced DDoS Protection application, which is a security service integrated with Cloudflare’s service platform.

Article Topics

case study  |  Cloudflare  |  edge computing  |  Edgegap  |  latency  |  online gaming  |  serverless