Smartphones that rival console performance are not a threat

ARM's ecosystem director Nizar Romdan told the Casual Connect conference in Amsterdam this week that mobile chipsets will reach graphical parity with the PS4 and Xbox One by the end of 2017. In some respects, it's a bold claim; Apple's ARM-based A9, currently by far the most powerful mobile chipset on the market, is a long way behind the likes of the PS4, and while some aspects of mobile phone chipsets may well catch up to the current generation of consoles in the next two years or so, it's unlikely that they'll touch the performance level of the consoles overall - bearing in mind not only the advantages of discrete GPUs, but the fact that consoles sport far more, and far faster, RAM than mobile devices do right now.

Still, I don't doubt that Romdan is speaking from a well-informed position - ARM's roadmap unquestionably has some components on it which will achieve performance parity with home consoles in late 2017. It's worth unpacking what that actually means. Leaving aside the fact that a few console-class components doesn't necessarily translate into a console-class gaming experience (the system has to be considered as a whole, and mobile phones are fundamentally not optimised for that kind of performance), it's also worth noting that for ARM to have these components on a roadmap for late 2017 probably means they'll be in consumers' hands in 2018 - while the consoles they'll rival launched in late 2013. The implication is that the time taken for performance levels in home consoles to be matched in mobile devices will have dropped to around four and a half years.

That's actually a really interesting achievement, albeit one so adorned with caveats that it looks like it's wearing a festival outfit made of ifs, buts and maybes. It's worth recalling that consoles aren't actually state of the art when they launch - PS4 was already outclassed by top-end PCs on the day that it appeared on the market, because it's designed as a good, sustainable trade-off between performance and cost. On a technological basis, comparing the very bleeding edge of mobile phone chipsets to game consoles, the reliable mid-range family sedan of the gaming world, is a bit of a soft target to choose.

"It's not impossible to imagine a future where your smartphone is your primary processing device, connecting itself to your keyboard and monitor to do work, or to your TV and gamepad for play"

On a business basis, though, it's still a really interesting comparison, because it implies that something genuinely new has the potential to happen in the next few years. For the first time since the arrival of smart devices, there may be phones on sale which match the performance of a top-end home console, within the lifespan of that console generation. That's never happened before; today's top-end smartphones are just about feeling out the margins of the performance of the Xbox 360 and PS3, several years after those systems were replaced by their successors. PS4 and Xbox One, if ARM's predictions hold true, will face equivalently powerful smartphones long before PS5 and Xbox Two (?) are ready to take the stage.

What does that mean? For many, it's going to conjure up once again the vision of a console-less future, one where smartphones have replaced the box under the TV entirely, and I have no doubt that the continued rapid progress of smartphones is going to provoke plenty more "death of consoles" narratives. That's just one possible future, though, and it's worth bearing in mind that while the lag between a performance level being achieved on a home console and that performance level being possible on smartphones will narrow further, it will never close to zero; in fact, much of the low-hanging fruit of smartphone chipset design has already been snagged, with fabrication processes reaching the 14nm scale in Apple's A9 and not a whole lot of potential for further shrinking beyond that without some serious advances (which would also, of course, boost desktop processors). It's fairly straightforward physics to say that desktop (or console) class processors will always lead smartphone chipsets in performance by a certain span of time; it's equally straightforward to say that a device plugged into a power source (and hence optimised for performance) is always going to have an advantage over a device with a battery (hence optimised for power consumption).

Still, let's do the thought experiment. Consider a world where the smartphone is as powerful, or as close to parity as makes no odds to the average consumer, as a home console might be. It's probably also pretty damned close in power to your average consumer's laptop, at that stage. Already, there's something slightly ridiculous about carrying around a device as powerful as your smartphone all day, but leaving it almost entirely unutilised while you work on your laptop and game on your console; it's not impossible to imagine a future where your smartphone is your primary processing device, connecting itself to your keyboard and monitor to do work, or to your TV and gamepad for play, shifting its role in your life as you move between physical contexts. Microsoft, in particular, has done interesting work on this front, experimenting with allowing its Windows 10 smartphones to act as desktop PCs when plugged in to a keyboard and monitor; it works rather better than you might expect and, in a world where smartphones really are that powerful (and most of your data stored in the cloud anyway, presumably), it's not hard to imagine this being a paradigm that would become quite dominant. In that case, the smartphone replacing the console - by becoming a console in its own right as soon as you put it in a room with a TV and a joypad - is also reasonably easy to imagine.

That's based on the assumption, though, that it's wasteful to have so much processing power duplicated between the devices you own - that having an ultra-powerful mobile CPU/GPU sitting on your sofa and being used to idly check Twitter during load delays in your game is wasteful and uneconomic, as is having your smartphone quietly charging on your desk during the day while you work on a laptop little more powerful than it. Yet it's equally likely that that's just what our future looks like - one where processing power is so incredibly commoditised and readily available that everyone owns and interacts with loads of powerful, under-utilised CPUs in their daily lives. Isn't this, after all, what has largely sunk the ship of "cloud gaming" - the realisation that there's so much cheap CPU and GPU power out there in the world that it's economically insane to try and centralise it all in a server farm, instead sucking dry the relatively scarce and valuable resource that is bandwidth? It's also worth bearing in mind that while the trend towards technological convergence is an extremely powerful one (I do like those side-by-side comparisons pictures which show that every piece of technology on a large, well-appointed desk in the 1970s or 80s is now integrated into a small smartphone), it runs into a brick wall at the point where consumers actually demand compartmentalisation. Build a camera into my phone? Great, I don't want to carry a camera and a phone anyway. Make me use the same device for professional work, for play, for family activities and for highly private moments with the Internet's less-clothed denizens? Yeah, quite a lot of people are going to take a rain-check on that one.

"The availability of a smartphone as powerful as a PS4 won't really matter very much to the PS4... They are devices that exist in very different contexts with very different paradigms of usage and interaction"

It's equally likely, in other words, that the availability of a smartphone as powerful as a PS4 won't really matter very much to the PS4. They are devices that exist in very different contexts with very different paradigms of usage and interaction; the fact that one is hypothetically capable of running software of equal quality to the other is something of great interest to engineers, programmers and people who argue with each other on technology subreddits, but not really worthy of a flicker of interest from consumers (who will consider it equally compelling information as "you know the engine in your food processor could also drive a pretty decent RC helicopter"). The rapidly falling cost of hugely powerful CPUs and GPUs will simply give us an embarrassment of processing riches; powerful game consoles, powerful phones, powerful tablets and powerful laptops, and while there will be crossover between their functionality and usage cases (I mean, who hasn't used an RC helicopter to slice an onion when the food processor is on the fritz?), their distinct presentation and context will mean consumers see no problem with having all of them in their lives.

Smaller, faster, more energy efficient and more powerful processors continue to open up exciting new possibilities and opportunities for games and for technology in general (I'm particularly keen on the idea of VR headsets being decoupled from cumbersome cables by the integration of smartphone-class GPUs, though I'm wary that it's only a matter of time before someone exploring a virtual world tumbles over a balcony here in the real and worryingly gravity-afflicted one). It's a leap of logic, though, to assume that developments in smartphone chipsets represent an existential threat to game consoles, or to gaming PCs; context is king, and of all the devices smartphones may eventually replace, other high-powered processing devices that fill very different niches in consumers' lives are perhaps the least likely of all.