This last week marked a rather interesting announcement that most of us (myself included, initially) probably missed.
Google this week announced the final realization of their Project Stream, now called Google Stadia. Stadia is designed to be a game-streaming service, which would allow you to play games on a remote, cloud-based PC, and stream the game to any device, where you would interact with a controller or keyboard and mouse as if you were playing the game locally.
If you’ve kept up with this type of technology over the years, well, you know that game streaming kind of…sucks. From startups that burned out, like Onlive, to nascent services like Playstation Now and GeForce Now, the common thread has been that the technology kind of runs poorly. In order to stream a game meaningfully, you need a lot of bandwidth – you have to be able to feed input data to the cloud system, which has to then process that input and provide the visual feedback, in high-definition, at least 30 times a second. If you want proper action gameplay, that needs to happen 60 times a second, and the newest craze in gaming, high refresh rate, is basically impossible.
The current platforms using this technology largely are limited by this. Playstation Now works okay, but action games tend to be poorly playable on the service. GeForce Now works worse and is still in a closed beta with a waiting list. So you can see why someone like me would be…not interested in such an announcement.
Into this market wades Google Stadia, and well, initially, my thoughts were the same. It’s just game streaming, so who cares?
Well, Google’s ambitions are perhaps a bit larger than just that.
To be clear, in the current game design space, there are a lot of games that can work reasonably well in such a service. RPGs, lower-speed action games, strategy games, many free-to-play style games – all of these run well enough and at a pace that befits such a setup. In such cases, the game quality would largely be limited by your own bandwidth – a fast enough connection would be needed to manage the downstream traffic (slightly larger than streaming HD from YouTube of Netflix) but consistent.
The problem many game streaming services encounter, however, is on the service side. Streaming a game is tough work, requiring a powerful machine to run the games on, large amounts of bandwidth, and a datacenter infrastructure that has enough nodes to deliver service close to you, so that the physical distance the data has to travel is not an obstacle. This is the chief obstacle for new competitors in the space – nVidia has only 15 data centers total for GeForce Now, which is something they are working to expand by using their new GeForce Now Alliance program, partnering with network hosts and internet providers to co-host their servers, and Playstation Now is only available in limited markets based on datacenter availability.
Google has a few advantages here. They own more servers and datacenters than anyone in the world (maybe Microsoft is close or slightly higher) and they own the interconnects and internet infrastructure that connects these data centers, with 7,500 plus edge nodes that reach every corner of the planet. They also have a daunting hardware and software reach, with Android as the top mobile OS in the world, Chrome as the top browser in the world, Chromecast as one of the top streaming devices in use worldwide, and a variety of available Chromebook minimalist laptops in the wild with low cost.
The other thing that Google is offering the others aren’t is a sense of platform. The technology is not solely intended to be a platform for existing games to be dumped onto, but is being demoed and intended to be its own ecosystem. Google has in-house developers, headed by former Ubisoft creative Jade Raymond, working specifically on games for Stadia. While some demos were of existing games (Assassin’s Creed, NBA 2K, etc) the exciting prospects were the new projects. One of the more interesting demos was a split-screen demo, where each player was playing cooperatively and could see each other’s screens. Another cool technology they demoed was the ability to state share, sending a link to your friend in a message which they could click and resume the game immediately from your state – allowing you to arrange simple score contests or to pass the baton off between friends and play a game together in a new way.
Some of the best parts are the fundamentals though – Google is pitching that they can deliver 4k, 120 FPS gameplay through the service, with a goal of getting to 8k over time. They’ve built custom servers with impressive specs and are even promoting the possibility of multi-GPU support to allow a premium tier with upgraded graphical fidelity. The service can stream gameplay (to you) and also stream your gameplay live to YouTube at 4k/60 FPS, allowing you an easy means to stream, as you would need no local hardware short of a device to play the streamed game on, which can be anything (a cheap, old laptop, an Android smartphone, a Chromebook, a Smart TV), although it wasn’t clear if you’d be able to commentate (I would assume yes, but couldn’t find much on that aspect).
These aspects give me a lot of excitement for Stadia – while I would not want to play a traditional game through such a service, stuff made for it, which leverages the advantages of such a service while minimizing the problems like latency, would be fun to play. The possibilities of such a gameplay mode would be great. Imagine an MMO where you could character trade with your friends, so if you have a full party and someone else wants to tank, you could swap characters and just play each other’s characters through a streaming mechanism, allowing you to send a link to trade gameplay instances. Or maybe a strategy game where you could invite friends to co-command your army, or could trade-off control of an army in a battle that is waged over days rather than minutes, allowing you to step into and out of the combat when it suits you.
The scale of gameplay can be changed as well. One of the examples Google used is taking the current Battle Royale trend and extrapolating it to over one thousand players, rather than 100. Stadia could enable this by keeping game servers in the same data center network as the streaming service, reducing the game server latency and allowing for more to happen. The implications it could have for other games is huge, too. Imagine an MMO hosted and played through Stadia – where there is one effective server with everyone on it, millions of players living in a single virtual world that can be persistent and changeable.
So while I think that the base offering of Stadia is just a more competent version of a thing I don’t want, it is in the flexibility and the platform vision that I find an actual desire to see where it goes. At launch this year, I don’t think it will immediately be worthwhile for those of us with strong gaming PCs.
But there is a future where the effects that such a platform unlocks makes using it far more worthwhile, and that is worth being excited about, to me.
(There’s also a lot of troubling implications for monetization, charging for services and games, and the pervasiveness of ads on such a platform – which I will probably explore in a future post!)