Google Stadia chief engineer Madj Bakar said that in a year or two, the game streaming system created under his leadership will be able to provide better performance and better response times compared to conventional gaming computers, no matter how powerful they are. At the core of the technology that will provide an incredible cloud gaming environment are AI algorithms that predict player actions.
The engineer made such an ambitious statement in an interview with the British Edge Magazine. Boasting of Stadia's achievements in implementing simulation and machine learning algorithms, he suggested that Google Stadia will become the benchmark for gaming performance within the next couple of years. “We think that in a year or two, games that run in the cloud will run faster and provide better responsiveness than running on a local system, regardless of its power,” Maj Bakar said.
As the engineer further explained, this will be achieved through proprietary streaming technology, which has already been tested as part of the Stream project. According to Google, the chosen approach will solve all the problems that one way or another arise in game streaming services due to the remoteness of data centers from the end user. The technology is based on “negative lag”, which should compensate for the delay that occurs due to the transfer of data from the player to the server and back. This negative delay will be provided by a buffer formed by rendering and transmitting “future” frames based on predicting the player’s actions.
In other words, Google Stadia’s artificial intelligence will try to predict what the player will decide to do at each moment in time and transmit to the player a video stream generated taking into account his expected reaction. That is, to put it simply, the artificial intelligence of Stadia will play for the user, and the user will see on his local device not the answer to his reaction, but the result of the game of the artificial intelligence, which went a little further than him.
All this sounds quite scary, but the first testers who have already tested the technology in action do not note any glaring oddities or inconsistencies. The full-scale launch of the Google Stadia cloud streaming service is scheduled for November this year, and then we will be able to evaluate how well negative lag works in real conditions. By the way, Google also plans to use adaptive screen frequency synchronization in Stadia so that users of its service feel as comfortable as possible.
Source: 3dnews.ru