When it comes to online gaming, users want immediate responses. Here's how Google Stadia has created what it calls negative latency.
Dan Patterson, CNET and CBS News Senior Producer, spoke with Majd Bakar, Google Stadia VP of Engineering, about the company's biggest challenge: Zero latency. Stadia, which was publicly released on November 19, 2019, is Google's cloud gaming service. This interview, conducted October 2, 2019, is the last of a three-part series. The following is an edited transcript of their conversation.
Majd Bakar: The biggest challenge is really to understand the network topology and how you can deliver the streams in the most efficient way. Gamers, they care a lot about the smoothness of the game, how the images are displayed on their screen, but also about the reactivity. If I press the button on my controller or on my keyboard, the action needs to take effect immediately.
Creating that tight loop between the input that the user is using and how we respond to it and display it to the user is one of the biggest challenges, and we think we solved that. We created what we call a negative latency. It allows us to create a buffer for the game to be able to react with the user, and that accommodates the latency that you have on the network.
SEE: Google Cloud Platform: An insider's guide (free PDF) (TechRepublic)
The only thing we couldn't solve is the speed of light. We still have to send electrons from our data center to your screen, and that is at the speed of light. In order for us to account for that distance, we go and create additional buffers for latency that the game would be able to recover. So effectively, you end up with an identical experience, whether you are running on a console or the PC under your desk or on Stadia.
The first thing you need to do is really take risks. You want to, because there's a lot of trial and error, and there would be a lot of failures. We've been working on this for the last four years; this is not something that we started six months ago. And, definitely, we had our shares of failures.
We also learned from the pioneer in the industry. OnLive was the first pioneer in cloud streaming, and they launched in 2010, and we learned a lot from their experience. As a CTO, you need to say, 'I'm willing to take risks, and I'm willing to try new things, and if some of them fail, well I learned from that. It's not a bad thing to fail.' This is actually something at Google we take very seriously. You learn always from your failure.
The other thing is, it goes--again--to iteration. You need to build a very quick iteration loop where you can experiment with things, capture data, and learn from this data, and iterate, improve your process. We did this at a very large scale last year with Project Stream where we launched a trial in the US for a very large number of users with our partner, Ubisoft and Assassin's Creed Odyssey, the game. We got a lot of data. We learned a lot about the Wi-Fi environment. We learned a lot about networking in the US. We learned a lot about how you can get that efficiency in the distribution of high-end games, and that is helping guide our design and our development for Stadia.
- Multicloud: A cheat sheet (TechRepublic)
- Hybrid cloud: A guide for IT pros (TechRepublic download)
- Serverless computing: A guide for IT leaders (TechRepublic Premium)
- Public cloud providers' network performance wildly varies (ZDNet)
- Best cloud services for small businesses (CNET)
- Microsoft Office vs Google Docs Suite vs LibreOffice (Download.com)
- Cloud computing: More must-read coverage (TechRepublic on Flipboard)