Google and Microsoft have recently announced that they are entering the world of cloud gaming, and the gaming populace reacted with…
“Really? Are we still doing this?”
Most cloud gaming platforms have seen only moderate success at best. Even Sony’s PlayStation Now platform is one of its least utilized services. So what is a cloud gaming service and why do they always seem to come up short? Let’s take a look.
What is cloud gaming?
When you play videogames, most of the time, you make use of a specific video game device. A PC, a smartphone, a game console, a handheld, all of these devices are essentially computers that run your game as software. It operates essentially the same way every video game back to the experimental 1970s prototypes did. You give it some input, it processes that input, and it returns the results of that input to the screen. From Spacewar! to God of War, all video games are the same.
Of course, this system has limitations. A game device can only ever play games compatible with its hardware. You won’t see an NES playing God of War, it’s just not powerful enough. This might seem painfully obvious, but it’s these exact limitations that have kept many casual fans out of the PC gaming world. The idea of building a PC with specific specs in order to run specific games is just too intimidating.
Cloud gaming was an idea that was supposed to remove these limitations by utilizing the internet and streaming video. The idea was simple, offload all the processing. Instead of having a device that was essentially a computer running a game, you could instead have a device that was a glorified terminal. Instead of running the game, all your terminal would have to do is show the game to the player. All the processing, graphics rendering, and input processing is done by the game running on a high powered server computer. The only thing your terminal has to do is display the graphics feed as streaming video.
This opened up a lot of possibilities. It meant that even low powered devices could run games as well as the most powerful PC. You could play the latest Call of Duty on your smartphone if you wanted. You could play Devil May Cry 5 on your TV without even owning a console. It was a great idea… on paper.
The problems with cloud gaming
You see, the promise of cloud gaming is allowing a server to do all the heavy work for you while all you have is a screen and an internet connection. This still means you have to send all your button inputs over the internet to the server, and then after the server processes those inputs it has to resend the game’s video feed back to you. That takes time because unfortunately, the internet has limits.
We have previously written about how even a perfect internet connection running at the actual speed of light will introduce a couple frames of lag when connecting up to a server half a world away. You can’t break the speed of light, so our internet connections will literally never get any better than this.
So the only way a cloud gaming service can reliably send near lagless gameplay to homes all over the world is to have server farms all over the world, local to all gamers using the service. In addition, these gamers will have to use an extremely fast internet connection and these server farms have to utilize the same high speed connection for uploading video feeds. All of this amounts to money, a lot of money, spend on both ends by both the consumer and the gamer.
And if the idea was to make cloud gaming easier by reducing anxiety over specs, it doesn’t really accomplish that if you have to have some sort of minimum internet speed to take advantage of it. That’s just a different sort of spec and it causes the same amount of anxiety in casual gamers.
This article started by reflecting on the failures of cloud gaming’s past. The very first cloud gaming demonstration happened in the year 2000 at E3. It was a small demonstration by G-cluster which streamed PC games to handheld devices over Wi-Fi and even though the streaming computer and screen were both local to each other, it still introduced quite a bit of input latency to the formula.
In March 2010 OnLive became one of the first publicly accessible remote cloud gaming services and it had a mixed reception at best. Few were willing to put money into its microconsole and most games were still plagued with heavy latency. In 2015, Sony acquired OnLive and converted this technology, along with the technology of another cloud service, Gaikai, to create PlayStation Now.
PlayStation Now is currently the world’s most popular cloud gaming service, but it has its issues. First, it still requires powerful hardware, which kind of defeats the whole purpose of cloud gaming to begin with. Second, even with the best internet connections, wired and local, PlayStation Now averages about 5 frames of delay, which for many pro gamers is unplayable. This isn’t so bad on 30FPS games, but once again you could just as easily play a 30FPS game on your local hardware. Add to this the $100 a year price tag, and you can see why PSNow hasn’t really caught on.
Google’s new streaming service, Project Stream, professes to use all new technology to reduce input latency, but unfortunately it doesn’t seem to be there. Reports have shown that it too averages about 4-5 frames of lag, comparable to PSNow, which honestly makes sense. Our internet infrastructure hasn’t greatly advanced since the debut of PSNow, so Project Stream would have to work with similar speeds. Both PSNow and Project Stream also top off at 30FPS and 900P resolution, giving them far worse graphical fidelity than even weak gaming rigs these days.
So what does Project Stream offer? Well it offers the convenience that PSNow doesn’t. It runs in a Chrome browser and is agnostic of platform. So you can run it on a PC, Mac or even Linux machine. You can also just as easily run it on a tablet or mobile device and switch between devices nearly effortlessly. Project Stream also automatically detects whether or not your internet and device is up to snuff for game streaming, reducing a little bit of that spec anxiety.
The same holds true for controller support which seems to be largely plug and play. Any Xinput or DirectInput controller simply plugs in and maps itself. PS4 controllers and Xbox One controllers just need to be plugged in to be used the same as if they were on a console.
And what of Microsoft’s new cloud gaming initiative? Well we have even less data on it than Project Stream. Microsoft is promising connections at 10 megabits a second and 54 worldwide datacenters, but without practical testing we cannot be sure what this will mean for input latency. If they actually do manage to drop latency down beneath four frames, it might just open up their streaming service to mainstream consumers, however we are at a loss as to how they could do that with our existing datalines.
Even then, this would only be the speeds that wired connections experienced. Wi-Fi connections are even worse, and there’s no real way to make that better outside of having individual users upgrade their own home network. PSNow currently averages around eight frames of delay on wireless networks, which once again is nearly unplayable.
So why now? Well for precisely the reason that no one has seemed to be able to perfect the technology yet. Since so few people are using cloud gaming services, it’s essentially an untapped market. The technology doesn’t have to be perfect, but if some company can find the right combination of low price and decent play experience, they would essentially jump-start the market with themselves at the head. This is very enticing especially with major game companies like Ubisoft professing that cloud gaming is the future of gaming as we know it.
The dangers of cloud gaming
While the idea of a gaming future with no consoles, no pcs, nothing but screens and internet connections is somewhat alluring, cloud gaming has a lot of dangers associated with it. The most notable danger is historical preservation.
Right now, we are having a very hard time historically preserving gameplay experiences. Any game with an online component cannot be preserved in a meaningful way. Retro games, while emulatable, are constantly under fire from companies like Nintendo making emulators and ROMs less accessible.
Many have pointed toward cloud gaming as a major bonus for gaming historians. Thousands upon thousands of retro games can be emulated from a remote server and played without needing to unearth hardware that can be up to three decades old.
However, some have voiced a more skeptical opinion, especially in adopting cloud gaming as the norm. Cloud gaming servers will only exist as long as they can remain profitable. Once they go down, they take everyone’s game collection with them. If we stream all our games we won’t have any way to preserve the code of these games and play them at a later date. In fact, we wouldn’t be owning games at all. We would simply be paying a fee to experience them.
And this is an attitude that the gaming world is rapidly gravitating toward. Games are no longer products, they are experiences that we pay to have close to their launch and then never again. Every new piece of online only DRM, every new exclusive digital distribution platform, every act taken against modders and homebrewers is a step toward giving up control of the very games we play to the people who make them. Cloud gaming may just be the inevitable end of that attitude.
What do you think? Is cloud gaming a worthwhile technology? Is it even possible with our current internet infrastructure? Would you pay for any of these new cloud gaming services? Let us know in the comments.