What is latency?

Users expect a fast response time to their commands when gaming online, but also when doing simple tasks on the computer. But low latency doesn’t always work — long distances and other hardware and software problems can slow down the flow of data. There are, however, concrete measures you can take to reduce latency.

Latency, also known as response time, describes the time between a command and the expected response. There is always a delay caused by a wide variety of factors. Basically, a low latency or delay is desirable — a high latency should be avoided.

What sounds abstract at first has a concrete meaning for most of us, es­pe­cially in tele­com­mu­nic­a­tions: When working on the computer, a program should start quickly after a double-click, when surfing a website should load im­me­di­ately, or when cloud gaming the decisive blow against the opponent should occur as close to real time as possible.

Es­pe­cially when using VR goggles, too much latency is also phys­ic­ally no­tice­able: The digital en­vir­on­ment does not reload fast enough, the realistic ex­per­i­ence and immersion is destroyed. In the future, low latency will play a much more decisive role in autonom­ous driving. Here, latency should be as close to 0 as possible so that cars can safely react to accidents and other sur­pris­ing events in road traffic.

In addition to Internet latency and computer or operating system latency, fiber optic latency — how long a signal takes to travel through the fiber optic cable — is also often important for everyday users. But there is also latency in other contexts, such as op­er­a­tion­al or mech­an­ic­al.

Even though the goal is to produce low latency as close to 0 as possible, in most cases this is not realistic. For example, Internet or computer latency is affected by the following factors:

• Distance between sender and receiver: For example, if you retrieve data packets from a server in Europe, the latency is sig­ni­fic­antly lower than if you retrieve data from US-based servers.
• Trans­mis­sion media: For example, there are different types of cable with different trans­mis­sion speeds. But also the type and number of in­ter­con­nec­ted hardware (server, hard disks, switch, ...) increases the latency.
• Packet size: The larger the data, the longer the trans­mis­sion takes.
• Type of data pro­cessing: Different com­pon­ents handle the pro­cessing of the data dif­fer­ently.
• Available bandwidth: The bandwidth available to you can slow down the trans­mis­sion. The signal strength can also have an impact, for example if it is par­tic­u­larly weak and/or needs to be amplified by a repeater.
• Queues at in­di­vidu­al stations, such as when an unusually large number of requests are made to a server sim­ul­tan­eously.

One of the most important factors for a good latency is a short distance between sender and receiver. At a higher level, this factor is processed via so-called edge computing, in which an IT ar­chi­tec­ture is set up that processes data as close to its source as possible. Beyond that, you can also make other op­tim­isa­tions to your hardware and software to improve latency:

1. When streaming or online gaming, make sure to select a server from Europe in the settings of the gaming or streaming platform.
2. Use a cable instead of WIFI for par­tic­u­larly data-intensive ap­plic­a­tions.
3. Reduce the so-called packet loss via your firewall. The firewall checks incoming packets and even blocks some — this costs time and sometimes data is lost. To prevent both, you can configure the firewall to specify which pages or ap­plic­a­tions do not require such a check.
4. Reduce the load on your con­nec­tion to have more bandwidth for the current ap­plic­a­tion: Close programs running in the back­ground on your computer, tem­por­ar­ily put other devices (smart­phone or tablet) in flight mode, and pri­or­it­ise your computer on the router. Un­ne­ces­sary programs are best com­pletely un­in­stalled.
5. Add more RAM to your computer.
6. De­frag­ment your hard disk.