Opinion: ON DX11 penetration, it's 'essential' nature, and the reality of why always expecting the best can be a bad and unreasonable thing.
“Wow consolised - another non DX11 game to make my high performance PC shake it's head in disgust.” – somemadcaant, commenting on a Witcher 2 gallery.
You know – and I mean no disrespect to the above poster – but it really is no wonder that there is a perception of PC gamers as a whiney minority. We can be guilty of it here at Atomic at times, too; the PC is our prime focus, and our most beloved gaming device. We often rate games using a similar attitude to features and graphical power, but a question occurs to us... is that really fair?
Now, sure, with a mission statement like MAXIMUM POWERED COMPUTING/GAMING, we can be forgiven for being a touch elitist – but hey, we’re at least upfront about it. But I also feel we’re beholden to step back from that every now and then to take stock of the reality of games development, and how the codemonkeys meet the needs of the gaming market.
So, about that DX11 thing... fair call, all the VGA industry is touting DX11 as better than sex. However, the cold reality of the situation is that the API only holds around 20 per cent market share, according to the Valve Hardware Survey. So that’s one in five gamers. On top of that, many of the bells and whistles of DX11, such as the much-hyped, little-used tessellation functions are so hardware intensive to leave even high end cards gasping for clock cycles. In our own testing, it’s pretty plain that you need a dual GPU setup to see these features running smoothly.
At time of writing, we can’t grab the SLI/Xfire stats on the Survey, but when we last looked it was a low single figure, which we can’t imagine has changed much. That means the 20 per cent DX11 share would be drastically reduced even further, likely down to a low single figure percentage.
While you’re considering the low number of gamers who can even take full advantage of DX11 features in a game, think also about the fact that developers are often a conservative bunch. The problem with developing to the highest spec is you end up with a game that no one can play – the Crysis conundrum. Unless you’re willing to develop a game purely as a large tech demo, it’s not good business sense.
And that’s even before we get into the added complexity of developing for the API. Sure, a lot of it’s very similar to DX10 before it, but it’s those new features that would suck up an inordinate amount of time – for the benefit, really, of not that many gamers.
Obviously, this has nothing to do with consoles. The Witcher 2, about which the above quote was made, is not even coming out on anything other than PC. We’ve had bit of a gripe in the latest issue about the just-released Shogun 2: Total War not supporting DX11 (like I said, we’re not immune), but when you think about it, why should it? For the majority of PC gamers, it’s simply not an issue.
And really, outside of the white elephant that is tessellation, can most of us even tell by sight the differences between DX10 and 11? I doubt I could. And there are still many other metrics of PC performance outside of that single signifier.
At the end of the day it remains our choice to be hardware enthusiasts, to ride that cutting edge of price and performance. But it’s always worth remembering that we’re not all there is to the PC gaming community. It’s churlish to consider that the industry must follow our lead above all others, not to mention that it would be highly damaging to our beloved industry.
Issue: 137 | June, 2012