"We, in no way, take credit for the idea."
LiquidSky CEO Ian McLoughlin knows video game streaming isn't a new concept. For years, various companies have promised players they'd be able to load up any game on any device via cloud streaming. Play the latest Fallout on an Android tablet or boot up the new Witcher at max settings on a four-year-old MacBook Air. It sounds too good to be true, and since the early 2010s, it has been.
LiquidSky is the latest company to promise low-latency game streaming. The premise is simple: Play any game you own on any Windows, Mac, Android or Linux device, no matter how outdated or powerless it may be. Every LiquidSky user gets access to a unique SkyComputer, where he or she can install new games, including titles like Overwatch and League of Legends, on a virtual high-end PC, or access their existing libraries from Steam, Battle.net, Origin and others. Play any of these games on any device at any time.
Once again, it all sounds too good to be true. McLoughlin is well aware of this fact. As a programmer and creator of LiquidSky, he's studied the turbulent, cash-sucking history of video game streaming services.
The most famous example is OnLive, a cloud gaming service that ended up $40 million in debt before finally disappearing in 2015. However, LiquidSky isn't just copying OnLive's model in a new technological era. McLoughlin thinks he knows where OnLive went wrong structurally -- and how to fix it.
"Those guys are pioneers," he says. "But there comes a time when you are pushing a rock up a hill, so to speak."
OnLive and other streaming services, such as Gaikai or Nvidia's GeForce Now, have traditionally used custom servers to handle each user's heavy lifting. This means the companies install physical servers at crucial locations around the world, and as demand increases, the number of servers also has to increase. It's a scalability problem, McLoughlin says.
Just look at what happened to OnLive: The company employed a 1:1 concurrency model, which essentially meant there was a series of desktop PCs in a warehouse each hosting one user at a time. Eventually, the company was able to host two or three players at a time, but it was an unsustainable system from the beginning.
"You have a million users flood in, you buy all these servers with massive capital up front, and those users are in different locations. There's too much latency, and the only games you can play are Lego Batman and Lego Star Wars," McLoughlin explains. "So you're left with this massive data center that you can't do anything with, so they started essentially giving things away for free. Even then, they couldn't get the users to enjoy the catalog. It was too soon before its time."
McLoughlin attempts to solve this problem with software rather than more hardware. LiquidSky has partnered with IBM to take advantage of the growing public-cloud ecosystem. All of the data center sites listed on LiquidSky's website are actually IBM locations, allowing the company to scale in real time at a relatively low cost.
For example, 40,000 people in Turkey recently attempted to access LiquidSky at the same time, and the nearest IBM bare metal server cloud automatically responded to handle the demand, McLoughlin says.
"We knew we weren't going to be able to take custom hardware and scale it around the world to get close to our users," he says. "But this big shift happened in the past 10 years in terms of public clouds, and now you can actually use public clouds like Amazon and IBM and the Microsoft Azure Cloud to do high-performance computing, which is enough to get to the point of running a game. That's sort of where we started building this technology."
On top of advances in cloud computing, the streaming industry itself has ballooned in recent years; services like Netflix and YouTube have fundamentally changed the way everyday people consume content. Streaming is now commonplace, which means service providers have had to improve their own systems to keep up with demand.
"Them having so many users accessing video content just forced the networking companies to upgrade their routers to stop all the loss from happening," McLoughlin says.
These factors -- the availability of public clouds and vastly improved upload speeds in homes -- provide fertile ground for a service such as LiquidSky to take root. Plus, McLoughlin has figured out a way to make LiquidSky free to access, something that existing services like PlayStation Now or GeForce Now don't offer.
McLoughlin is banking on 2017 to be the year of cloud gaming. It looks like the global technological infrastructure is finally ready to support high-quality, low-latency video game streaming, and major names like Samsung and Sun Microsystems co-founder Scott McNealy are throwing their money behind LiquidSky. McLoughlin courted these investors, in part, by proving LiquidSky can work in the real world.
Over the past two years he's tested out his streaming infrastructure on roughly 1.2 million devices, covering everything from Android smartphones to gaming desktops. That's another big issue facing streaming service providers -- within the PC space, the sheer number of unique devices (and related bugs) is staggering.
When developers build games for the Xbox One or PlayStation 4, they know the precise specs of each console and can focus on tailoring their games to the hardware. This isn't the case with PC development. There are countless custom PC configurations in the world, and they can each produce unique glitches. Even McLoughlin's own MSI gaming laptop crashes when he attempts to boot up Battlefield 1 through his SkyComputer, even though his rig can definitely play that game easily on its own.
Device fragmentation is one reason LiquidSky has been in beta testing for two years. In late February, the service will finally start rolling out to the public and the true cloud-streaming trial will begin.
If all goes well, LiquidSky won't stay contained in the gaming world. McLoughlin dreams of unifying the technology industry through cloud computing. Just like he's attempting to make every Windows, Mac, Linux and Android device run the same games in the same way, he eventually wants to make cross-platform functionality standard across all industries.
"We want to build a place where you can go in and just click something and it opens," McLoughlin says. "It doesn't matter whether it was designed for Android or Windows or Linux or a smart TV; you just click it and it opens. And then you go to a different device, you log into the Sky or it's already logged in, and you're right where you left off in that application. That's sort of how we see this evolving."
But, for now, the future starts with gaming.