2016 was a big year for virtual reality devs: muliple headsets became commercially available, and with them came new markets that are filling up (though perhaps not as quickly as you might like) with potential customers. So what's next?

According to Valve chief Gabe Newell, the future of VR is all power: lighter, more capable headsets that are hooked up to PCs much more powerful than what most people own right now.

"VR is going to drive much higher computational and GPU loads than we see out of existing desktop games," Newell said recently. 

"Like, I would argue that the smartest thing AMD and Intel and Nvidia could do is give away VR development systems to software developers, because we're going to consume the shit out of Intel and AMD CPUs! Much more so than high-end games are doing right now."

Plus, he's confident that PC-powered room-scale VR will progress to "house-scale" VR in the near future, as it becomes possible to knit several "room-scale" VR spaces together into a single space. Of course, you'd need room-scale-capable wireless VR headsets too, but Newell believes they're a "solved problem" that are right around the corner.

"My expectation is that [wireless] will be an add-on in 2017, and then it will be an integrated feature in 2018," said Newell, speaking during a recent press briefing attended by Gamasutra and other outlets. The conversations that ensued saw Newell and fellow staffers opening up a bit about the state of the company's VR hardware and software development efforts.

More notably, for devs who are making VR games and want to continue making VR games for the foreseeable future, they spoke a bit about where they see the VR market going in the next few years -- and what steps the company is taking to try and nudge it forward.

As we reported yesterday, Newell says Valve is currently developing three new "full" VR games, using both its own Source 2 engine and Unity. These are games Valve expects to charge for, games the company hopes will attract a lot of attention.

Designing VR games with "mass appeal"

"We're definitely trying to build stuff with mass appeal," Newell says, as opposed to the sort of prototypes and experiments (like The Lab) that have dominated Valve's VR software output to date. According to Newell, the folks at Valve have learned enough about VR development that they're willing to start making some bigger bets.

"We have this theory: we think we can make three big games, we think that we know enough now to do that, and we're going to find out if that's the case. We're pretty sure that all the game developers are going to learn positive or negative lessons from what we do, which is sort of where we think we need to be," he said.

"There will be really clear design choices that we're making, in each of these games. They're very different. Each game developer will be able to look at those and say, that was great, that was not as great. Which is part of, from our point of view, that's a useful charateristic of these three."

Pushing VR game design forward is important to the folks at Valve because, according to Newell, there are some great VR experiences out there -- but not enough to justify millions of people buying PC-driven headsets. In Valve's eyes it's quality/quantity of available software, rather than price of entry, that's keeping a lot of people from buying into the VR market.

"If you took the existing [PC-driven] VR systems and made them 80 percent cheaper, there's still not a huge market. There's still not a really incredibly compelling reason for people to spend 20 hours a day in VR," said Newell.

"We actually think that if anything, most of the interesting stuff is going to be happening on the high end. That we're actually resolution-constrained, we're CPU-constrained, we're GPU-constrained. That's where the interesting stuff is gonna happen, not on the low end of the market. Once you've got it, once you've got something and you can say okay, this is the thing that causes millions of people to be excited about it, then you start worrying about cost-reducing." 

"There's sort of an old joke that premature cost reduction is the root of all evil," he added.

Building better VR hardware, informed by the needs of VR game makers

Valve is also working to improve its VR hardware in conjunction with its VR game development efforts, a symbiotic arrangement that Newell likens to the way Nintendo handles game and hardware design.

"One of the questions you might ask us, why in the world are we making hardware?" Newell said. "What we can do now is we can be designing hardware at the same time that we're designing software. This is something that Miyamoto has always had, right. He's had the ability to think about what input devices and the design of systems should be like, while he's trying to design games. And our sense is that that's going to allow us to actually build much better entertainment experiences for people."

This has been the goal from the start, according to Newell. When it comes to the business of building VR hardware, Valve sees it as a valuable complement to its games and services businesses, rather than as an end in itself.

"The idea isn't oh, we've suddenly thought we could make more money by building hardware," said Newell. "Hardware's actually traditionally been kind of a lousy, low-margin business. But in order to take what we think are interesting steps forward in terms of the kinds of experiences we design, we need the ability to be thinking more about everything the customer is doing."

So how has that actually worked out, now that Valve has had Vive hardware on store shelves for nearly a year -- and been experimenting internally for far longer? 

According to Newell, devs can see an example of how this process pays off by looking at the mockups of Valve's new VR controllers, which surfaced at Steam Dev Days last year and are far more akin to Oculus' Touch controllers than the wand controllers which come with the HTC Vive. 

A (poor) snapshot of a prototype of Valve's new VR controller, which is meant to rest in the hand

These new controllers "came from doing the first generation [of Vive controllers]; working on our games, people working on the games were like 'here's stuff we would actually use.' And then they'll mock it up, we'll do a hardware prototype, they'll go back and change the game. The process is iterative," said Newell.

"And I think we have our own internal development process, and then we take it out to our development partners, other game developers, and say 'so this is what our thinking is.' So Robin Walker will sit down and say, 'this is why we think this is important for what we're developing. And this is how we express that in terms of the next generation of hardware design.' Then we ask other people, 'does that match what you're thinking, or not?' And we get feedback from that."

This iterative process of making VR games, designing VR hardware around those projects, then taking it outside the company for feedback, seems to drum up a lot of excitement among Newell and his colleagues. It feels a bit like the way Oculus execs were talking about "hand presence" (via their Oculus Touch controllers, naturally) back in 2015 -- there's this sense of anticipation, like game makers will soon have a whole new, mature and multi-faceted method of input to explore, one distinct from gamepads, keyboards, and mice.

"It feels like we've been stuck with mouse and keyboard for a really long time," Newell said. "And the opportunity to build much more interesting kinds of experiences for players were there; we just needed to sort of expand what we can do. But [for us] it's not about being in hardware. It's about building better games. About taking bigger leaps."

Hoping VR is more like the DS than the Wii

However, the folks at Valve are quick to point out that they may totally drop the ball -- that they are, in fact, trying to get comfortable with the notion that VR as a whole might still fail in spectacular fashion. Newell says he's optimistic about the future of VR, but admits he's been wrong before.

"You can always be surprised, right. Like personally, I think the DS was kind of stupid, right? I was totally wrong. I thought Sony was going to crush Nintendo in that generation of handheld devices," said Newell.

"I hadn't worked on it, I hadn't tried to design any games for it, and clearly the DS ended up being the winner. The flipside was, the first time I played Wii Sports -- just to continue using Nintendo as an example -- I was like oh my god, there's so much opportunity, there's so much potential here, that we're all going to go discover. And then it sort of turned out that Wii Sports pretty much nailed it. And that was it! Everything else just ended up....there was less sort of innovation, to mine, in that direction. So even when you're doing it, even when you have a really good instance where it seems like there's potential, you can still find out there isn't much there."

He went on to acknowledge that the VR market is still volatile, and it's yet unclear if/when it will see a significant expansion.

We certainly saw the effects of that volatility last year, when some studios released VR games that became breakout successes -- Survios' Raw Data, Owlchemy Labs' Job Simulator, and Neat Corp.'s Budget Cuts spring to mind -- and many others released VR games that were "wildly unprofitable."

"Yeah, the people who built VR, its sort of this trade-off," said Newell. "You say 'okay, how much risk am I taking on, and how much of an opportunity am I giving myself. How long is the runway gonna have to be before I really start to see explosive growth.' And each developer has to decide that."

Going forward, Newell thinks most VR devs will continue to make VR games, the volatility of the market be damned. "I'm pretty sure there are almost no developers who are working on VR titles who are like saying 'Oh, I'm going to back off,'" he asserted. "They're all coming back and saying 'Oh, I'm gonna do more VR!'"

However, he wants VR devs to appreciate and manage the risks that still plague the nascent VR game industry, and he reiterates that Valve is committed to working with developers to help them sort out funding concerns -- without any expectation of an exclusivity arrangement.

"One conversation we have with some developers is around how they manage their risk, right. It's like you've got people building proprietary walled gardens who say be exclusive to us and we'll give you this bunch of money. And we're like, we hate exclusives. We think it's bad for everybody, certainly in the medium- to long-term, and I'd probably argue in the short-term as well," said Newell. 

"But we're happy to say to people, look, you need to figure out how to manage your risk so you can develop the title you want to build. So let's have a conversation where we can help you manage your cashflow over the course of the development so you can go and build the thing you want. And get it out to market and start getting your money from where you should be, which is from your customers, rather than somebody else. We're perfectly willing to help people do that."

"And then they say well, does that mean we have to be exclusive to Steam, or exclusive to the Vive, and we say oh hell no," he continued. "You guys should be making the best decisions for your customers, not having somebody else steer that. That's when we have conversations about that, it's usually around 'how do we pay the bills while we're building our next VR title.' And that's usually a pretty productive conversation. But it's really about smoothing out risk over the lifetime of VR investments."

As we approach the one-year anniversary of the HTC Vive's debut, it's clear that Valve continues to view VR as less of a new platform (like, say, Steam Machines) and more of a new way of computing -- one that it's still intent on pushing forward. 

The company is making a show of doubling down on VR game development, iterating on its VR hardware (devs should also expect an improved version of its Lighthouse tracking technology in the near future), and assuaging devs' fears about surviving and thriving in the fledgling VR market.

"There's that little bit of tension between some people going out and saying oh there'll be millions of these, and we're like, wow....I don't think so," said Newell. 

"Because I can't point to a single piece of content that would cause millions of people to justify changing their home computing at all. It's like a great thing for enthusiasts and hardcore people, and where we are today is way further down the road than we were a year ago, but it's just going to be this slow, kind of painful, fits and starts kind of thing."