“It's a real game changer for us. This allows for a console-like experience on the PC." Valve head Gabe Newell’s proclamation at the 2011 Consumer Electronics Show earlier this year was bold, demonstrating the developer’s confidence in Intel’s new processor architecture, Sandy Bridge.
His enthusiasm is well founded. PC gaming has long been the preserve of the dedicated, and though Facebook has attracted vast new audiences to the idea of playing games on their PCs, a great many machines are incapable of running anything more graphically demanding. And for the uninitiated, the world of gaming PC components, from GPUs to motherboards, isn’t only confusing but also off-putting, making the consoles’ comparative lack of barrier between player and game rather more enticing. In short, the PC is failing to live up to its enormous – and growing – potential as a game machine.
But Sandy Bridge could change all that. While Intel’s chipsets have long offered integrated graphics, Sandy Bridge boasts a significantly more powerful GPU on the same silicon die as its CPU. The new design is both faster and more power efficient, allows for more extreme overclocking than has been previously achieved on Intel chipsets, and represents a significantly more consistent lower target for PC developers to shoot for.
Valve's forthcoming Portal 2
“There are tonnes of advantages: the performance that you're getting out of it; the price that you're getting it for; and the heat that it's causing and whatnot,” Valve vice president of marketing Doug Lombardi tells us, explaining Newell’s stance. “On the processing side, it’s allowing a lot more cycles without charging the customer more for fans and cooling or the part itself. Then on the GPU side, it's bringing that baseline integrated graphics up quite a bit, which several people have been very vocal criticising Intel about over the past several years. With Sandy Bridge, they're going to get a part that's probably going to run everything that's out there now. [Intel] demoed WOW on it, which is fairly intense and obviously fairly popular and we're running Portal 2 on it.”
The upshot of this integration is the very real prospect of a future in which PC specifications are more standardised - at least at the lower end of the market. Not only would this mean delivering consistent gaming experiences will be less difficult and expensive for studios, but also that gamers daunted by the impenetrable and fragmented world of PC gaming may finally see a way in.
For those who already consider themselves avid PC gamers, however, the term ‘integrated graphics’ has become a dirty one. With capabilities far below those of bespoke graphics cards, the gulf seemed insurmountable. “While we were making Trials 2, it was a real eye-opener to find that many PC owners aren’t actually hardcore gamers with really high-end desktop PCs and expensive graphics cards.” admits Sebastian Aaltonen, lead programmer at developer RedLynx. “We had to implement these low-end modes with simple graphics and simple lighting - but we actually got it running on Atom netbooks! Sandy Bridge is finally bringing Intel close to Nvidia and ATI’s graphics performance, so you don't have so much code difference between chips.”
RedLynx CTO, Joonas Tamminen (above, left), and lead programmer, Sebastian Aaltonen
In hindsight, such streamlining seems obvious; why haven’t we seen a base-level standard that can run today’s, and indeed tomorrow’s, PC games in a hassle-free environment before now? Is it a question of cost, or simply that the market for capable but accessible gaming PCs wasn’t recognised? Intel’s director of gaming strategy, Randy Stude, reveals that Sandy Bridge has long been on the table: “We had it on the road map four years ago. We couldn't talk about it then, it's just a trend whose time has come.”
But if the PC is heading towards a consistent and highly capable base-level standard, might this lead gamers who cut their teeth on eminently accessible social titles to venture a little further into the world of PC gaming?
“I think that people who aren’t buying a gaming PC deliberately might at some point notice that they have a really good gaming system due to these new processor types,” thinks RedLynx CTO Joonas Tamminen. “So that might be a segment that could be lured into gaming because more and more people will start to have machines that can run games with good framerates at a high quality without investing in a specialist setup.”“What's happened is that weird range of computers that we have to ensure [a game] worked on, just got [much smaller],” adds Lombardi. “So you just chopped off that whole bottom edge. It's not going to disappear overnight, but 18 months from now you're not going to have to worry about that guy sitting on their on DX8 anymore, 'cause somewhere in the past 18 months he probably spent $500 to get a new machine - thank god we don't have to work with DX8! [Laughs] We can now spend our time on other things rather than optimising for the people who haven't bought a new machine or, god help them, did buy a new machine but don't know what a GPU is.”
Stude concurs, before recounting some eye-opening figures on the PC gaming market. “We see this massive audience of gamers playing on the PC that's growing exponentially year-over-year; EA has said that it thinks there’s something like 290 million folks playing games on Facebook alone. According to the research we helped sponsor as part of our membership with the PC Gaming Alliance, there's something in the order of 350 million gamers playing on their PCs worldwide. Clearly a very minor number of those would fall into the enthusiast category so there's a much larger mainstream audience.”
Intel’s technology doesn’t spell the end for the likes of Nvidia and ATI, though, and a healthy market for bespoke, high-end graphics cards will no doubt continue to exist for some time yet. Indeed, no one we speak to can imagine a world without them – at least, not in the near future. But by closing the gap between the liquid-cooled, LED illuminated monsters that hum in rows at so many LAN parties and the humble budget laptop, Intel has empowered the entry level users of an already bottom-heavy market.
Valve vice president of marketing, Doug Lombardi (above, left), and company head, Gabe Newell
But could the need to cater to an ever-expanding new audience divert PC development’s attention away from the pomp and splendour that has cemented its place at the top of the graphics food chain for so long?
“[Developers are] already doing that in a sense, we do select the target segments we want to go for,” thinks Jaakko Haapasalo, producer at 3D, PC and mobile performance benchmarker, Futuremark. He goes on to explain that while the huge number of possible setups PC developers have had to support has made such targeting difficult, Sandy Bridge’s known quantity could see certain types of game focus solely on the base-level. But, he stresses that this won’t be the case for all developers, and that upgrading from DX9 to DX10 hardly represents a revolution.
Stude is similarly pragmatic: “There's just too many risks venturing high on the code. One thing we've learned over the last several years is that going high on your graphics or even your processor expectations does well for you from a marketing perspective, because you have plenty of excitement from us on the processor side and the graphics vendors want to help promote your game. But that doesn’t ensure market success. Examples you might look at are games that overreached perhaps too much, like the first generation Crysis game. I was right there saying we'll happily support all the processor enabling [Crytek] want because we expected great things from that title - and it was a success by any measure in terms of its sales - but I don't think it lived up to the expectations of the market or the developer, probably because it over-reached the majority of the PCs out there. And you can see with the shift in the subsequent titles, WarFace and Crysis 2, they're dipping much, much lower in terms of the system expectations to reach a much broader audience.”
Crytek's Crysis 2
But what of Gabe Newell’s hopes for a “console-like experience”? Stude prefers to think of it as stability, pointing out that this is the first time that developers will have a capable base-level specification which they can be confident players will have access to.
Aaltonen is more effusive: “When these new chips really decrease the fragmentation of PC hardware so it's much easier to create PC titles that really work and require less support, we can get a much broader audience to run the games that we have. If in the future we decide to do an Xbox 360 game for PC, it would be much easier because the low-end hardware is so much closer to Xbox 360.”
“Game development is sort of on a plateau right now,” adds Stude, “where the capabilities of the leading console platforms typically dictate the performance required for the majority of [PC] games that are shipped to market – especially in the west. How many games are really taking advantage of DX11? Not too many. Why not? Well, because the consoles don't support anywhere near that capability. But when the consoles catch up in their next cycle, my personal belief is that a reinvigorated code base for games will start to support them and they'll find their way onto all platforms. There's no such thing as a PC-only title necessarily - maybe a few extra widgets at the behest of relationships with graphics companies, but nothing extraordinary that you would call market momentum right now.”Very high-end PCs currently sport 16GB of RAM and twin (or even triple) graphics cards, but very few games take advantage of such headroom. It therefore may seem that we’re reaching a point of diminishing returns, even while there is still plenty of distance left to run when it comes to graphical fidelity (see our recent feature on PC graphics to learn more about the current state of that sector). And the growing majority of gamers are satisfied with less demanding software: Haapasalo doesn’t think Sandy Bridge will do much to whet gamers’ appetites for more.
“Frankly I don't see any new revolutionary capabilities in absolute terms. We're still going to have animation, AI, physics, graphics... I don't think this particular generation has anything revolutionary to offer,” he says. “The direction of integration is interesting, it may enable some kind of efficiency that we haven't seen yet in the future, but I don't see a quantum leap - at least at this point. But bringing up the lower end is interesting. And it's not just about performance: we tend to think a lot about the API generations and for us the move to a DX10-capable, integrated and possibly very widely spread device is even more interesting than raw processing power or other considerations.”
The first Sandy Bridge processors hit the market earlier this year but Intel rapidly ran into problems when a fault was identified with the Series 6 ‘Cougar Point’ chipset used in every system. The issue lay with the chip used to control Cougar Point’s SATA II ports, which Intel discovered could physically degrade over time. Despite Intel’s own calculations that the fault would only affect five to 15 per cent (depending on usage) of customers over a three-year period, the company issued a recall of the eight million chipsets it had shipped that hadn’t already made it into consumer’s hands.
Due to the long-term nature of the problem, and the low percentage of those who will likely be affected, Intel advised early adopters that they could continue using their second-generation Core i5 and i7 processors while it worked out a permanent fix. A revised chipset which fixes the issue has already gone into production, but these units won’t reach OEMs till the end of February while full production isn't expected until April. The cost of replacing the faulty chips, along with the lost revenue from processor sales, is likely to cost Intel around US$1 billion. It’s hardly an auspicious start for PC gaming’s great new hope.
The delay also brings Sandy Bridge into closer proximity with major competition in the form of AMD’s second-generation Fusion processors. The first of these, Ontario and Zacate (which are intended for netbook use), are already available, while more powerful variants will follow in the next few months.
Lost sales notwithstanding, it seems unlikely that Intel will suffer any long-term damage to its reputation from the recall thanks to the swift manner in which it handled the problem. But equally, the vast majority of customers who buy a Sandy Bridge PC are unlikely to have any relationship with the company beyond their recognition of the logo (and jingle) from all those TV adverts.
Futuremark producer, Jaakko Haapasalo (above, left), and Intel director of gaming strategy, Randy Stude
As such, it’ll be fascinating to watch as, in the coming months, the revised systems start to make it to market and developers begin to demonstrate what’s possible on the new architecture. “I guess people maybe don't realise, but there's an absolutely huge install base of PCs that don’t have high-end discreet cards,” muses Haapasalo when we ask him what the future holds for Sandy Bridge. “If that base is going to be brought up even a bit by this, then I think it's going to have an impact. If a large section of those gamers then suddenly jump to being able to play fairly high-end 3D content, that is interesting, but whether there's going to be the right marketing channels and the right business models for that to happen, I don't know."
Either way, whether you term it ‘console-like’ or ‘stable’, it’s clear that Sandy Bridge, at the very least, will change expectations for what PC gaming can, and should be.