So far there’s been a lot of stuff said about the new generation of consoles, whether it’s the Xbox One isn’t as powerful as the PS4 or even if the next generation as a whole isn’t much better than the previous generation.
A lot has been said on the subject of optimisation, this is the process which – in laymens terms, programmers change the code so it works better for a certain set of conditions. Now when the various experts on this topic, you know the people making the games, the people who made the consoles etc, said that the first bunch of games for the next generation aren’t as optimised as they can be, the obviously more knowledgable random forum posters picked up on the fact that both of the new consoles use a set of hardware that is largely the same as a home PC. So, they posited, you don’t need to learn how to optimise it – you already know how to develop for the PC.
The strength of a home console, the main reason it manages to even slightly compete with a home PC whilst still costing a fraction of the price is that it has a very static set of specifications. Optimising code for a PC is exceptionally hard, because there are literally thousands of potential builds out there. The GPU alone can vary massively in power, then you have the CPU, the speed of the hard drive and eventually you even have to worry about the operating system (various versions of Windows, Linux and Mac).
To make money from the PC market a game has to be able to run on the largest set of specifications possible. This means that an average PC game is pretty much a badly optimised pile of crap with about a thousand options to try and get the most from whatever set of components you attempt to run it on. Now, this doesn’t matter for the most part. Even though the software is relatively unoptimised (note, when I say badly optimised – it really is relative, games developers are the top of the tree at this stuff – even the most badly made professional game is still 100% more optimised than, say, Excel) an average gamer PC has so much graphics bandwidth that it can afford to push more pixels than it really needs to, you’ll still get exceptional performance.
The console market, on the other hand, is a sitting target. You know beforehand exactly what you need to aim at. You know the GPU speed, you know the CPU speed. For the most part you know how the hard drive is going to perform (as long as the user follows Sony’s guidelines if they replace their drive). This means you can take a device that is several times less powerful than the PC and still be able to get the same performance.
Unfortunately, it takes a while to know exactly how to get the best from what is still an incredibly complex device. Plus there are always little tricks you can do to squeeze more out. But it takes time. This is the reason why games seem to get better and better through a consoles lifetime, developers learn to use the hardware more efficiently and learn more tricks to wring the power from it.
If you’ve been here since the last generation, you probably remember how it went. The 360 came out first. By the time the PS3 came out the developers were producing some pretty amazing titles on the 360. The PS3, however, started getting a lot of really bad ports. Slow frame rates, worse textures. This wasn’t because the PS3 was any less powerful, it really wasn’t and time has proven that. This was because Microsoft got the advantage of being there over a year earlier (from the time the dev kits went out) so developers got to grips with the 360 first and every PS3 game ended up being a pretty unoptimised port.
Fast forward to today and that’s exactly what we’re seeing now. Both consoles are experiencing almost straight ports from normal PC based hardware with little or no optimisation, but the One is being hit the hardest because Sony managed to turn the tables and get the PS4 dev kits into developers hands early. So the One is experiencing games like Ghosts, BF4 and AC4 which have received some PS4 love but very little One love. Game reviewers have jumped on these differences as a sign of the lower ability of the One. In reality the difference is probably about 9% if you’re talking about literal bandwidth differences – the 50% number came from Edge and it’s a total fabrication, the numbers just don’t add up and the ‘developers who we interviewed’ developers seem to consist of the same indie developers they interviewed the week before who are currently working on Sony exclusives and at least one of them stated they hadn’t even asked for an Xbox dev kit. Also they claimed that the dual memory swapping system the One uses was harder to use than the PS4 or the 360 before it. Except it’s the exact same system the 360 used, except with faster memory. Several developers i’ve talked to have stated it took no time to get used to, because it was pretty much identical.
But the point is, don’t judge the next generation on this first bunch of games. They are largely unoptimised and pushed out the door quickly for a launch. Especially don’t trust Ghosts which is a terrible, terrible bit of software and appears to be a heavily patched version of the same engine IW have been using since the original IW left. Seriously Activision, stop screwing your customers for more profits – hire an actual engine development team and re-write the CoD engine from scratch. While you’re at it get some people with actual ideas in there too.
So, wait for the next set of games – this generation is going to be pretty amazing. Killzone and Forza show off what the platforms are truly capable of graphically and using those you should see what’s going to come.