Google Stadia

Google Stadia ‘Negative Latency’ Technique Isn’t Actually New for Gaming

  • Twitter
  • Facebook
  • Google+
  • Pinterest

The hype around video game streaming is ramping up as we get closer to actual launch dates. Currently, there are a couple of serious competitors in the market, Project xCloud and Google Stadia. However, only the latter is anywhere close to actual release, launching next month across the world. Google is also probably miles ahead of everyone else in terms of infrastructure required for game streaming.

Despite all the marketing claims and infrastructural superiority, people are still skeptical about Stadia. There are still massive hurdles that just cannot be crossed by having more data centers. One of these is the inevitably higher latency and input lag while streaming. There is no way an input signal can travel to the data center and back faster than it does on your local PC or Console. However, Google has a few tricks up its sleeve to solve that.

Google claims it can achieve ‘negative latency’ on Stadia

Apparently, Google has found a way to overcome the latency and input lag problems on Stadia. According to a Google exec, games on Stadia will actually be able to respond faster than a local Console or PC. This is an astonishingly bold claim from Google but is such a thing even possible? Well, according to Google, yes.

Google Stadia
Stadia Data Centers

However, things are not exactly that straight forward. Google will use two different techniques to minimize input lag and potentially get the ‘negative latency’ effect on Stadia. The first one is increasing the FPS of the games to the point where the latency between input and the action in-game is automatically reduced. This is pretty obvious as usually higher framerates lead to better response times.

The real difference, however, will be the second technique. Stadia will use predictive input in games to predict what the next move of the player will be and then match it with actual input for faster processing and ultimately, a reduction in input lag. So, the ‘negative latency’ is actually just a marketing term for input prediction. However, people are not exactly happy about the technique.

Don’t be mad about predictive input in Google Stadia

As soon as the news broke out about Google using predictive input in games to predetermine a player’s actions for reduction in input lag, people got mad at Google for trying to ‘ruin competitive multiplayer’. They claimed that the game predicting their next move would completely ruin the accuracy of the gameplay and would result in bizarre player movements and decisions in game. However, what people don’t realise is that we already use predictive techniques in most multiplayer games.

Prediction has been in games for ages

Branch prediction is a widely used trick in programming to bypass processing lag. Video games have been using the technique for a very long time now, especially in multiplayer. Remember when multiplayer video games used to be really laggy and choppy? Why isn’t that the case anymore? While our connections did get better, video games also managed to get really good at hiding lag.

Call of Duty Modern Warfare

Call of Duty is a prime example of prediction in games. Sometimes, the events on your game don’t correspond exactly with the events in a replay. This is because the game actually guesses what your next move would be and feeds it to the server. The server then decides and syncs it up with what is actually happening on the server, which is why there are bound to be some discrepancies here and there. This is also why games look much smoother these days even during connection and data interruptions.

Stadia uses the same technique, but in reverse

Google Stadia is doing the exact same thing, but in reverse. In Stadia’s case, the server predicts what your next move would be, aligns that with your actual input to produce a viable result and then sends it back to your client, which is just a video decoder here. Furthermore, it would be easier for Google to provide a more uniform experience since most of the processing from all players will be happening on the server itself.

In addition to input prediction, Google Stadia is also using predictive modelling to generate frames in advance and only present the ones that would appear in front of the player based on their actions. This would obviously take a lot more bandwidth but Google can definitely handle that with its engineering and infrastructure capabilities.

Doom Eternal

Google Stadia launches in 14 countries in November. However, Doom Eternal, which was supposed to be Google Stadia’s big launch title in November has been delayed till March 2020. Would you be checking Stadia out on launch now without Doom Eternal?

Leave a Reply

Your email address will not be published.
Required fields are marked *

error: Content is protected !!