With the PS4 Pro and Xbox One X firmly established in-market, thoughts have already begun to turn towards what sort of console might be coming in the not-too-distant future. The PlayStation 4 lineup isn’t very old by console standards — it’ll be five this year — but one report suggests Sony is already sampling dev kits for the upcoming PlayStation 5.
That’s the word from SemiAccurate, but most of their story is behind a paywall. It’s been summarized elsewhere (whether it’s been summarized accurately or not is impossible to determine) and it doesn’t give the kind of specifications one might hope for. The PS5 would use AMD’s upcoming Navi architecture as its baseline and be built around a custom Zen microprocessor, according to Resetera. Dev kits have supposedly been seeded in such numbers, the SemiAccurate author believes a 2018 launch isn’t out of the question, and there are supposedly VR “goodies” baked in at the silicon level.
Of all the ideas contained herein, the one genuinely odd one would be launching a console in later 2018. First, it’s odd that Sony might consider this step given that we’ve gotten no publicity — at this point in the PS4’s hype cycle, we already knew the console was coming later in 2013, even if we didn’t know much else.
The case against the PS5 launching so soon is also easy to understand. When the PS4 debuted in 2013, one of the chief complaints from gamers was that it didn’t offer enough of a visual upgrade compared with the PS3, despite the seven-year difference between them. Jumping for the PS5 now, just two years after the launch of the PS4 Pro (See on Amazon), would trigger the same kinds of complaints. That’s particularly true since we haven’t even seen a single major process node shrink since the PS4 Pro debuted. Node shrinks deliver cooler operating temperatures and lower power consumption, which is the principle reason why the Xbox One X (See on Amazon) and PS4 Pro were able to pack in dramatically better performance without driving noise, thermals, and power consumption into the stratosphere.
While it’s likely true that AMD’s Zen can deliver better performance, than the company’s low-power Jaguar architecture, Zen is still a larger core than Jaguar was. Again, this is where process node shrinks come in handy, but the modest move to a refined 16/14nm node (12nm in marketing speak) likely doesn’t give Sony all that much additional headroom. 2019 or 2020 seems like a much more plausible target for a major console launch.
Backwards Compatibility Would Change the Game
The only way it makes sense for Sony to launch a new major iteration of its game console this quickly is if it intends to retain full backwards compatibility. There’s no reason why it couldn’t. Unlike previous console generations, which often made radical leaps from one architecture to the next, maintaining x86 compatibility for an x86 application isn’t a very big deal. Similarly, maintaining GPU compatibility between the Polaris-class GPU inside the PS4 Pro and a Navi-class GPU inside a hypothetical PS5 shouldn’t be a major issue, either.
The larger question would be whether Sony wanted to follow the so-called “iPhone model” when it comes to maintaining backwards compatibility between console generations. It’s one thing to say that the PS4 must play every game the PS4 Pro plays, and something else to say the PS5 and PS4 Pro must play the same games. Gamers who knew the PS4 Pro and PS4 maintained compatibility would probably rightly expect that to continue to be the case. Another alternative is maintaining backwards compatibility across the PS4 / PS5 generation, with the PS4 being targeted for lower resolutions and reduced visual assets. This would also work, but inject still more complexity for developers and might slow down sales of the new platform, since gamers would presumably choose to stick with hardware they already own. All of this can be avoided if the PS5 can play PS4 games but not vice-versa, and that’s the route we’d expect Sony to take.
The alternative would see Sony leave people to buy a brand new console in the shortest window between launches in the modern generation, at a time when it’s going to face a challenging endeavor in delivering higher performance without using a larger chassis or higher-end power supply. And let’s face it — what reason does Sony have to try and steal a march on Microsoft? The PS4 has outsold the Xbox One by more than 2-to-1. First mover advantage doesn’t always work out that way — just ask Nintendo about the Wii U — and Sony doesn’t need to launch a quick Hail Mary to reestablish themselves at the top of the market.
We’re not saying that Sony couldn’t launch a new console this year or in 2019. But we’d be very surprised if it did without some form of backwards compatibility with the PS4 generation.