PS4.5 and Xbox One.5 Could Come With a Hidden Challenge for Developers
With the rumors about a PS4.5 or PS4K finding multiple supporters, and Phil Spencer’s long term plan for a faster hardware evolution for Xbox, the coming of a mid-generation iteration of the PS4 and Xbox One is a realistic possibility.
I personally don’t see it as something we should bet our monthly wages on just yet. Planning and prototyping is one thing, and actually putting a product on the market is another. Yet, I wouldn’t be excessively surprised to see it happen, given the signals, so it’s certainly interesting to discuss about it.
There’s one main challenge I see on the horizon for this kind of initiative, and it’s full compatibility for games between the original consoles and their newer, relatively more powerful iterations.
The most relevant and recent precedent is Nintendo’s 3DS, that evolved into the New 3DS. While it’s backward compatible, it’s not fully forward compatible: there are games exclusive for the New 3DS, leaving those who bought the original portable out unless they upgrade.
The plans outlined by Phil Spencer appear to involve full backward and forward compatibility, and and the same seems to emerge from the rumors about the PS4.5. While this is great, it also comes with a hidden challenge for developers.
You probably noticed that, despite their limited and rather cheap hardware components, current generation consoles are very good at running games that require much more powerful PCs to function with comparable performance. This is not because consoles are some sort of miracle machines built with salvaged alien technology, but simply because they come in a sealed box with a single configuration, which means that developers can optimize their games down to the pixel to work with that configuration and nothing else. The gains in terms of performance are pretty massive.
The introduction of more powerful, mid-generation consoles would involve the fact that developers would have to optimize their games for two configurations instead of one. This potentially mitigates the performance gains coming from having better hardware to begin with. The modular X86 architecture mitigates this problem, but doesn’t completely cancel it.
Of course, this is still easier than developing for PC. Two configurations are certainly more nimble to optimize than hundreds thousands, but we’ll have to see what the impact will be. Developers might be able to still squeeze similar gains, considering that that two hardware setups aren’t even close to the galaxy of different combinations of components you can found on PC, but it’ll definitely require more development and QA resources (configurations don’t just need to be optimized for, but also tested for).
Developers faced with the necessity to support two configurations can either pile in more development and QA resources to make sure that every configuration is best served and optimized, or they can put most of their resources into one of those configurations.
The latter option would involve either focusing on the higher-spec console, leaving the lower-spec one running on its own devices, without caring too much if it performs badly, or focusing on the lower-spec console, allowing the higher-spec one to just offer whatever gain the more powerful hardware involves, even if it’s not fully optimized, and doesn’t juice the components as much as it could.
I seriously doubt that focusing only on the newer iteration would find much purchase among developers and publishers. After all, it’s quite probable that the lower-spec configuration will remain the majority of the installed base for quite a while. Focusing mostly on the older iteration is something that I see as more probable, at least for smaller developers and publishers that don’t have the possibility to considerably increase the resources spent on their games.
This means that, at least for games that adopt this solution, it would take quite a while for the more powerful hardware to be fully supported and optimized for.
Of course, bigger publishers and developers would be able to eat the higher development costs, and work on both configuration to squeeze out their full potential, but whether they actually would remains a question mark. In the end, it’s always a matter of money and budgets, and we have to remember that those two configuration would potentially become four for games on both PS4 and Xbox One.
There’s also the possibility that some publishers would not increase their development and QA budgets, just stretching them over the two configurations and simply doing what they can, but of course that would lower the optimization and polish on both, potentially bringing to consoles a limited degree of the messy PC ports we have seen in the past years. Hopefully, that won’t happen.
As usual, the developers that would probably get the most out of the new hardware are the first parties (and potentially second parties as well), that receive the most support directly from Sony and Microsoft. When you have the console manufacturers eating the increased costs for you, optimizing for two different configurations becomes much less of an issue.
This leads to what’s probably the best solution. If Sony and Microsoft really want to push forward this idea of mid-generation iterations of their consoles, they’ll need to minimize the impact on third party developers as well.
One way to do this, is to configure their development APIs so that they do as much as the development work as possible for the developers. This, though, can prove problematic. The more an API allows developers to work “close to the metal,” the more it shifts the responsibility to make things work on the developer itself. This is especially true (at least to our knowledge) with DirectX 12, since Microsoft publicly stressed on this point several times. It’s not that the same doesn’t go for the PS4, but we know a less about its API, since Sony kept its evolution more under wraps.
Ultimately, Sony and Microsoft will probably have to help developers through the challenge of optimizing for two configurations one way or the other, either by developing tools for that purpose, by providing direct support, or by focusing their research and development efforts so that the two configurations are so similar that the differences in optimization aren’t very relevant. Of course, this means that the new consoles wouldn’t be a massive evolution from the older ones.
At the moment, most of this is pretty much up in the air, as we don’t know much about Sony and Microsoft’s specific plans. Yet, this is probably one of the most relevant and interesting challenges console manufacturers and developers will have to face. Time will tell, but we’re quite possibly in for an interesting ride.