In part two of our interview with Nate Mitchell, VP of Product at Oculus Rift, we discuss upcoming resolution improvements to bring the tech into HD, how the reaction has been from both indie and major developers, and how they’re dealing with simulation sickness issues.

Catch up with part one of our massive Oculus Rift interview here.

GON: You’ve mentioned that you’re focusing on one single kind of Rift for the consumer launch. Can you explain the reasoning behind that? Some might argue that there’s another niche of older, cashed-up hardcore gamers who would be happy to pay two grand for a really high-end version, on top of the consumer version? Isn’t it just a software issue, as the biggest difference would probably be resolution?

Nate Mitchell: There’s two main reasons. One is that we want game developers to only have to target one version. We want them to be able to say ‘ok this build of the game runs at 60 frames per second, at this one resolution, and it’s perfect for the Rift’. Having that is crucial for game developers; they don’t like developing for multiple specs. That’s one of the biggest problems with gaming on the PC, it’s so fragmented that there are all sorts of weird problems. That’s part of the beauty of the iPhone, the consoles – they’re single hardware specs.

GON: But PC devs are used to developing games that support multiple resolutions?

Nate Mitchell: That’s true, but the second point I want to make is that we haven’t sorted out all of the details there. There’s the whole question of how does the form factor of the Rift change with different LCD panels. There’s an incredibly amount of hardware decision making that goes into the Rift design using the panel that we use. It’s not like we can use the same exact panel, but at higher resolution. It’d require a full redesign from the ground up to go with a different panel. So for right now we need to stay laser focused on shipping one design, one Rift. In the future, if we turn into a billion dollar company manufacturing our own displays, I’d be happy to offer a deluxe version of the Rift with a higher def screen and all the bells and whistles, but these screens cost millions of dollars to develop. It’s out of our scope right now, we’re very much piggybacking off all of the display manufacturers, using their technology for our product.

GON: Do you think Oculus will ever develop and release input devices for VR, or are you  always going to stick entirely with the HMD?

Nate Mitchell: We are incredibly interested in human-computer interaction. There’s a couple of our guys here that focus all their R&D on researching and evaluating what the best input device for VR is. The dream, when you imagine VR, is being able to look down and see your hands, see your fingers and interact with the world as you would naturally. We’re not necessarily there today but there is Microsoft’s Kinect, and the Leap Motion, and a number of other companies in that space doing some incredibly interesting things. We really believe that the future of VR gaming as a platform will go in that direction. So the short answer is no comment at this time, the long answer is it’s a super-interesting topic and we’re constantly researching it.

GON: Is input a harder challenge than the headset challenge?

Nate Mitchell: I wouldn’t say it’s harder – I would say it’s different.

GON: On to the subject of simulator sickness, what are you doing to minimise this?

Nate Mitchell: Man, that could be a whole interview by itself. It’s impossible to talk about all of it. For the most part, we need a couple of things. We need positional tracking, that’s a major potential factor right now as we don’t have it yet. From there we’re going to continue to improve the SDK, adding things like hopefully positional tracking, more predictive tracking, head and neck modelling, all those tiny features make a huge difference. The final step is the content. We’re working with game developers endlessly – a huge portion of my GDC talk was specifically dedicated to what devs can do to avoid simulator sickness. We’re developing a best practices guide that outlines a bunch of this stuff. For example, right now, any positional tracking movements that you encourage in the game can create some sensory conflict. For example, having a lot of computer screens that look like the user is supposed to bend down to read. Also, moving the player backwards through space at high speeds seems to have a pretty negative effect on people. Taking away camera control for a cut-scene and removing head tracking is no good. Having the camera move in a way that isn’t related to head movement is a sure fire way of making someone feel simulator sick.

It’s a matter of educating developers – I was just talking to DICE yesterday and they’re super excited by the challenges. But they don’t see them as show stoppers, they see them as “how can we design the game to lend itself really well to VR”. We’ve already seen developers tap into this. The CCP guys developed EVR, this VIR space sim, and did this incredible job where they took all the things we’ve been talking about and applied them to their game. You’ll notice if you have a chance to play EVR – you can only move forward in space, all of the cut-scenes and UI are integrated into the game world, the player never loses control, there’s always head tracking. There’s even a great visual identity with the player’s body sitting in the cockpit. When you combine all that you get this really cohesive experience, and your brain says “wow, I’m really piloting a space ship”. It feels awesome.

GON: Big budget developers are really risk averse – they’re spending hundreds of millions of dollars so have to be sure their game will sell. Rift is still going to be quite niche for at least the next five or so years, do you foresee that most of your software support will come from the indie market? Or is it that these big budget guys are so incredibly excited about VR, you know personally passionate about it, that we’ll see a Rift mode sneak into the bigger titles? You were just at Dice, is that what you saw?

Nate Mitchell: It’s impossible to say. Let me just answer a couple of things that you said. First of all, we don’t believe that Rift is a niche product at all. In fact we believe that it has an incredible attractiveness to the mass market. It all depends on content. If you ship only Battlefield 3 on the Rift then yes, it’ll only attract the hardcore gamers. Which isn’t necessarily niche, but let’s say it is for argument’s sake. If you have Minecraft on the Rift, that’s a very different market segment that’s going to want to try Minecraft VR.

I certainly think that in the beginning, a lot of the excitement and new experiences will come straight from the indie community. We’re doing a lot to support the Indie community, with Unity integration, the SDK, and we’re going to continue to do what we can to make it as frictionless a platform as possible. Even if you’re not a developer, if you buy our dev kit you can develop something, share that with people and get involved with the magic of VR.

I think what you said though is absolutely true, the reason that Hawken is coming to Rift is that we walked in and showed them the Rift and they were like “Wow, you’ve done it – how can we use this with Hawken, this will be so cool”. We hear that over and over and over. Take John Carmack – John wasn’t thinking there’s money to be made in VR, he was thinking “this is incredible, I want to step inside Doom”. What we hope is there’s a couple of Rift evangelists at every company that push for, even unofficial, Rift integration. That’s all we need.

GON: Anything else you’d like to add?  

Nate Mitchell: Yeah, what do you think of the Dev kit?

GON: I’m a graphics whore by nature, and have tested a lot of VR headsets. It’s definitely the best I’ve seen by far, but resolution is a huge issue. Today’s games are so detailed due to HD screens, and all that detail gets lost in the low resolution panel. So games with simpler, sparser environments tend to look better. I’m not even sure 1080p will be enough – the immersion comes at the cost of clarity. It’ll be interesting to see if people find that trade off acceptable. I also found scale in games rather strange – proportions of landscapes and objects that look fine in 2D suddenly look really fake in 3D. So devs will need to start using more realistic scales and proportions.

Nate Mitchell: I promise you that we’ll be fixing that (resolution) in the near future. We’re also working with more developers so there should be more made-for-VR stuff in the near future. Two quick points. One is that Vireio Perception is not the way it’s meant to be used. We’re not against it, but we do stress that driver-based VR is not the best experience. Also, give us the benefit of the doubt – this first version is a tool for developers. When we do, if we do, show a higher resolution display, I’d be really excited to hear your thoughts. I think you will be impressed.

Nate Mitchell is the VP of Product behind the Oculus Rift, the virtual reality headset turning heads and stomachs with its impressive technology and affordable entry-level pricing. Bennett Ring sat down with Nate to talk all things VR, what’s in store for the Rift, and whether or not the PS4 will be powerful enough to run their tech.

GON: You’ve been associated with some interesting start-ups. What attracted you to Oculus?

Nate Mitchell: When we met Palmer I was obviously intrigued. Palmer is an incredibly interesting guy, an incredibly smart guy, and when we got talking about the Rift, it became clear there was this great blend between Palmer’s hardware experience and our software experience – Michael Antonov (Chief software architect at Oculus, formerly of Gaikai and Autodesk), Brendan Iribe (CEO of Oculus, formerly of Gaikai and Autodesk) and I . Then when Palmer actually showed us the Rift, just the most basic demo, I remember the moment I put it up to my eyes, and then took it off, I knew immediately that I wanted to be involved in the project, and that this was the future of gaming. At that moment we jumped on board and decided to invest our time and energy to make Oculus a next generation VR gaming platform.

GON: Looking at the eventual launch of the consumer version, what do you think will be more important – having perfect hardware, or nurturing a rich software ecosystem for it? And how do you balance those two very separate requirements?

Nate Mitchell: It’s a great question. It does really rely on both of those things being great. In the most basic sense, great software is a must-have. Without software, the hardware is really nothing. But our team’s focus is on hardware. The way we balance it is to not focus too much on the games. We have game developers out there making games, we’re not making first party content. Our goal is to build the absolute best hardware possible, giving game developers all the tools they need, and hopefully that in turn cultivates that great software ecosystem, that brings us the triple A, killer app for the Rift.

GON: There’s the open source Vireio driver that has been developed by fans of the Rift, why aren’t you doing that internally when it seems to be a really promising way getting a large library of software to work with your hardware for a minimal amount of effort?

Nate Mitchell: Vireio Perception was actually developed by a guy called CyberReality, a member of the MTBS3D forums, and we’ve since hired him as our community manager. We really, firmly don’t believe that a great VR experience can be achieved at the driver level. I think that if you play any of the games that Vireio supports, there are a handful that work pretty well. For example, Half Life 2, which is all first person, very few cut-scenes, basically all done through the eyes of Gordon Freeman, that’s a perfect fit, right? So Vireio Perception with Half Life 2 on the Rift is a decent experience, but the vast majority of the games that the driver supports, you get a mediocre, if not poor experience, that really disorients the user. You can’t have low latency head tracking when you’re doing mouse emulation. Even the best games just don’t play well in VR with Virieo Perception – or any other driver. It really does need to be implemented by the game developer.

That’s something we’ve preached since day one, because VR is hard. VR is incredibly challenging. The UI alone requires game design changes, let alone camera controls and things like that. That’s why we’re investing so much into creating a great SDK, to let developers really tap into the hardware and develop an ultra-low latency VR experience. Everything else, those are nice hacks, and they might be cool experiences, but it’s not something where we would invest our time. We’d rather focus on “made-for-VR” gaming, or ports done by the developer, at the metal, where you get the highest quality experience.

GON: Do you think there’s a lot of crossover for developers of first person games, so they’ll be better at VR?

Nate Mitchell: That’s really hard to say. It really depends on the developer and the experience they’ve created. First person lends itself very nicely to the Rift, but we already have a number of third person developers and RTS/God games investigating the Rift having as much success, if not more, than first person games.

GON: In terms of the consumer launch, from your perspective what’s the number one thing you need to do to make this the killer device?

Nate Mitchell: The number one thing on the top of our list, since day one, has been a higher resolution display. The Rift is all about immersing you visually in the virtual world, and the display is at the crux of the experience. Right now, the dev kit is still relatively low resolution because of the high magnification. Bringing that res up, hopefully in the 1080 range, increasing pixel density, just ups the level of immersion so much. It makes it a much more fun experience. We’ve been investigating that stuff in the office for a while. I can promise you that it helps quite a bit.

GON: I’ve been using the Rift dev kit, and the resolution is obviously a big issue. You mentioned Half Life 2, and when I was looking at Alex, the female companion, her head was just 10 pixels across. This is at relatively close range. You lose a lot of her emotions. It made me think that maybe 1080p probably isn’t going to be the ideal solution either. Do you think it’s realistic to aim for a 2560 x 1440 panel in the next year and a half?

Nate Mitchell: That’s hard to say, because I’m not deep in that display industry, but obviously that tech is there and people are working on it. Whether or not that’s something we can tap into I can’t say. But what I can say is that 1080p in a form factor like the Rift is approximately double the pixel density of the developer’s kit. It’s a night and day difference. What you’re saying is true, you really want a 4K display in there, but even that 1080p is such a huge jump in terms of immersion and visual quality that it’s certainly going to be enough for the consumer version of Rift. But yeah, you’re absolutely right, I’d love to see even higher resolution versions. There is another challenge though, as the game has to render a consistently higher resolution every time we up the resolution of the display. One subtle benefit of the current panel is that even high-end games can render at that low resolution and easily hit 60 frames per second. Jump to a 4K screen and you’re not going to render Battlefield 3 at 60 frames per second. Maybe you can, but you’re going to need some truly serious graphical horsepower to get there. Again, we’re just on the cusp of the hardware being ready for all this. You look out five years, and you’re right, that’s the direction we’ll go.

GON: It’s interesting that a lot of the media coverage of Rift doesn’t cover what’s powering it – the system at the end of the HDMI cable. They want higher res, without thinking about the PC requirements to power it. On that note, the Playstation 4’s specs are out, do you think it’s powerful enough to run a 1080p Rift?

Nate Mitchell: Absolutely. It really all depends on the content. Look at two random games. Minecraft running on a PS4 versus Battlefield 3 running on a PS4, there’s a completely different graphical horsepower need under the hood to get those running for the Rift. You’re going to be able to have quality high-end smaller experiences in the consumer version, running off a tablet. Even potentially Ouya or the Ouya 2. It’s not necessarily the hardware that is the bottleneck, it’s the content. I think you’ll find the PS4, based on the specs they released, has definitely enough horsepower there to get some high quality Rift experiences. It’s going to really depend on how game developers leverage the hardware.

Based on what I’ve seen of the next-gen engine technology, like Unreal Engine 4, people are designing their engines this time around for 60 frames per second up front, there’s now a stigma about 30 frames per second, which lends itself really well to transition to VR gaming.

GON: Sony has experimented with consumer Head Mounted Displays in the past with their HMZT1 and T2, are you concerned that they’re going to come in and copy the Rift’s idea, pumping their immense resources into their competing product?

Nate Mitchell: No. The Rift SDK is not a DLL that anyone can just hack into. The whole SDK and all the work we’ve done is tied to the Rift hardware. So Sony would have to copy our hardware…

GON: That’s what I’m saying though – how do you stop Sony or another company from looking at what you’re doing, how you use the lenses with the screen, and using that basic design to build their own VR headset? Do patents protect you enough to stop Sony from doing that?

Nate Mitchell: We are well protected and we take Intellectual Property protection super seriously. I can’t go into any of the details, but it’s not something that we’re worried about. I think what is a more realistic situation is one of the big game manufacturers like Sony, stepping up and saying they’re going to develop their own competitor to the Oculus Rift. Different SDK, different system, but same goal of consumer VR. At that point it’s just a question of who has the better hardware, who writes the better software. Everything we do here every day to build the Oculus platform, we’re confident that we’re going to stay in front. It’s a challenge.

GON: How do you stay in front when you’re facing off against these multi-billion dollar behemoths, who have thousands of engineers? What is it about Oculus that puts you in this rare position to be the leader in such an exciting new hardware field?

Nate Mitchell: I don’t want to give away all our cards, we’ve got a couple of things up our sleeve that give us an advantage. But I can say that people in very large companies are very surprised at just how much we’ve done in such a limited time frame. Even more than that, and I think it’s the crux of it, is the building of a great team with top talent in the industry to work on this next-gen platform. We’ve already started that process, we’ve built an incredible team, who we believe are the best in the world to deliver consumer virtual reality, and we continue to expand that team even today. Incredibly smart, passionate developers working on this thing, iterating in ways that big companies just can’t. We’re building our dream, hoping it comes together.

Tune in for part two of our massive Oculus Rift interview tomorrow.

The AMD Radeon HD 7990 is one of the strangest reviews I’ve had to write in quite some time. You see, the first dual-GPU Radeon HD 7990 was released “unofficially” by PowerColor way back in December of 2012. Then ASUS released the ARES II, another version of the 7990 living off a H2O diet thanks to its built-in water cooler. Yet it’s taken AMD several more months to get around to releasing the official version of the card.

Normally AMD comes first with a new product, then the others follow suit with their unique versions, so I’m most interested to see what happens when that procedure is turned on its head.

Both the PowerColor and ASUS cards were built on proprietary designs, but AMD’s 7990 is based on the new reference design, the template for all future 7990s. Unlike the other two cards, which could happily heat an igloo during a blizzard, AMD’s version of the 7990 is much cooler, with a TDP of just 375W — well down from the 500W seen on the first two. This dramatic drop in hot air blowing out the back of your case is possible because AMD can cherry pick the very best GPUs for its version of the 7990, and they’ve made sure their selected GPUs only need the slightest whiff of power to run.

It’s no surprise then that the huge triple fan cooler manages to keep everything cosy without sounding like a screaming cyborg baby, with basically whisper quiet operating volumes.

Strapped beneath the cooler are two Radeon HD 7970 GPUs, and they haven’t undergone the hobbling of the earlier Radeon HD 6990’s twin GPUs. All 8.6 billion transistors are present and accounted for, and they’re basically identical to the standalone 7970, with one slight difference; the maximum frequency is now 1GHz, rather than 1.05GHz. It’s a minor change, but it could result in slightly slower performance than a pair of Radeon HD 7970s in CrossFire.

There’s also a gargantuan 6GB of memory, but it operates as if it’s just 3GB while in action, due to the limitations of CrossFire (NVIDIA’s SLI has the same issue). Twin 8-pin power connectors supply the juice, and the entire package fits into a dual slot package, far smaller than the four slots required by twin graphics cards. At a hefty 12 inches it’s even longer than Dirk Diggler though, so you might have to move some hard drive bays around to squeeze it in.

If you just happen to have five monitors lying around the house – don’t we all? –the 7990 will happily power them all simultaneously. The four mini-DisplayPort and DL-DVI outputs at the rear of the card provide AMD’s excellent Eyefinity technology, the undisputed choice for multi-monitor mavens. When the card is not in use, AMD’s new Zerocore power technology can shut down one entire GPU, as well as the fans cooling it, helping to deliver a respectable idle power use of just 20W.

In case you hadn’t noticed, there’s a whole heap of power on this graphics card. The Radeon HD 7970 uses arguably the fastest affordable GPU on the planet right now (the NVIDIA GTX Titan doesn’t count, as its insane price means that only the world’s reclusive Trillionaires, living in secret bases on the moon, can afford it), and the 7990 has two of them. Before we look at how this behemoth runs though, I need to discuss a problem that has become a real issue for AMD over the last couple of months – stuttering.

As any AMD CrossFire user can tell you, playing games with twin AMD cards results in noticeable hitching or stutters, and it happens as often as every few seconds. Unfortunately it’s not detectable by our usual frames-per-second benchmarks, so NVIDIA came up with an incredibly complex procedure called FCAT that can measure the issue precisely, spitting out graphs that would confuse a mathematics professor. Obviously they didn’t do it out of the goodness of their hearts – they did it to show that their dual-GPU method (SLI) doesn’t suffer from the issue.

I don’t have the $2000 video capturing machine required to run FCAT yet, so I had to resort to a much simpler, yet no less useful, method of detecting stutters – I played games with the 7990. Gasp! I often don’t have time to play games properly with the cards I’m testing, but with stuttering being such a huge issue for AMD at the moment, it was the only way to detect if the 7990 had the same problem of CrossFire’d 7970s.

Unfortunately, stuttering was present in every game I played. As an NVIDIA SLI user, it’s a very annoying issue that I haven’t seen before, and it happens even when the frame rate is well into the hundreds. I’d go so far as to say that it’s a show-stopper of a problem until AMD fixes it, but luckily it looks like a solution is on the way. AMD supplied special drivers that attacked the problem, and it definitely helped the issue, if not quite cure it entirely.

Sadly, there’s no indication of how long it’ll take until these are released to the public, and it’s likely they’ll cause a performance decrease.

Now that I’ve got that nasty business out of the way, I can talk about the benchmark results. I ran the game in several recent games against its direct competitor, NVIDIA’s GeForce GTX 690, and the performance was overall slightly better than NVIDIA’s card. The lead was never huge though, with even the biggest gap being just 9%. Note also that these results used AMD’s latest public beta drivers, and not the special stutter-killer set, as per AMD’s requests.

I also didn’t have the hardware to measure performance against dual Radeon HD 7970s, but by all accounts the 7990 is about 10% slower. Considering it’s about $300 more expensive, you might be wondering who would bother with the 7990? For starters, it’s definitely quieter than twin 7970s in action, something that speaker users will value greatly. Secondly, it only needs two slots inside your case, rather than the four of a CrossFire user, which could be a major boon for those with cluttered cases. It also comes with eight kick ass games, which helps to lessen the blow somewhat.

Finally, it is possible to run twin 7990’s in CrossFire, but you’ll probably need a small fusion reactor to power it.

If AMD could iron out the frame stutters, I’d have no qualms in recommending this card to those who want the fastest single-card, dual-GPU solution on the planet. Unfortunately these stutters come with AMD’s latest public drivers, and they’re ruining the gaming experience. So instead might I suggest a case of wait and see; if AMD resolves the issue entirely – and games.on.net will tell you if and when they do – only then will it be the time to go and blow your credit limit. It should only take a month or two before we know one way or the other, which can feel like aeons to somebody wanting to upgrade, but in this case patience is the most prudent option.

Good:

• Strong performance, edging out NVIDIA’s 690 in most games
• Low TDP
• Very quiet
• Only takes up two slots in your case

Bad:

• Super-annoying stuttering bug is still present (though hopefully solved soon)
• Slower and more expensive than twin 7970s

At a recent BioShock Infinite showing, games.on.net had the chance to sit down with Irrational’s Bill Gardner, User Experience Specialist on the upcoming game. We spoke to him about all things PC, including Irrational’s dedicated PC support team, how DX11 is improving their experience, and the difficulties in making characters like Elizabeth interact well with the player. Take a look below, or download the video in HD from our file library.

A YouTube mirror is also available.

I’ve got a couple of disclaimers to make before I wax lyrical about Dead Space 3. Firstly, I’m a massive fanboy of the series, with the original making it into my top 5 of all time. Secondly, I’m a bit of a horror aficionado; if a movie doesn’t have cats jumping out of closets, ghosts revealed in mirrors or a shower-curtain reveal scene, it won’t make it into my collection of 300 horror DVDs and Blu-rays.

Now that I’ve got that out of the way, I can explain why I think Dead Space 3 is easily the best in the series, despite the fact that the last third of the game won’t have you reaching for the Valium quite as much as its predecessors. Any horror fan can tell you that making a sequel to a horror film is hard; making the third in the series scary is almost impossible. Horror of the edge-of-your-seat, hold-your-partner-tight kind uses a relatively small set of tricks, and reusing them is a sure fire way to bore your audience.

Yet, the first two thirds of Dead Space 3 use very similar techniques to the earlier games. You’re still traipsing around derelict space ships, using the dynamic lighting of your weapon to light up the dark corners infested with all manner of deformed nasties. Except now — and this is the masterstroke of brilliance — you’re doing it with a pal.

You might think playing with another person lessens the tension, and that’s absolutely true if your co-op buddy is a douchebag who thinks teabagging Necromorphs is hilarious. Even the most perfect horror ambience can’t withstand such dickish behaviour. But if you find a co-op pal (shout outs to Dr3money, who watched my back) who treats this experience the same way a good horror movie should be approached – in silence, without cracking lame jokes every five minutes – you’ll find it to be even more emotionally exhausting than playing alone.

Now you’ve got a buddy to look out for as well as yourself, and all too often you’ll spot an axe-wielding zombie about to hack his/her head off from behind way before he/she does. Approaching death arenas with a pal, the large open areas where you just know you’re gonna get swamped by the ‘Morphs, usually involves several minutes of discussion about how best to handle it, followed by several minutes of getting your face ripped off as you watch your strategy go down the crapper. Co-op is a godsend, and makes the rehashing of scary tricks that much more enjoyable. It’s also integrated perfectly, with cut-scenes reworked for the co-op game, and even co-op specific missions.

If there’s one disappointment with co-op, it’s that the promised player-specific hallucinations are very rare, feeling more like a token effort than a game-changing feature.

Dead Space has always been known for its fantastic environments, and the third delivers even better game spaces. Each area tells as much of a story as the audio and text logs scattered throughout, and I often found myself browsing the photos and personal items in people’s rooms and homes. The last third of the 20 hour campaign takes the player down to the icy planet of Tau Volantis, a refreshing break from metallic corridors, and introduces one of the many new novelty features, body heat. It’s also the area where the game morphs from tense corridor crawler to more of an all-out firefight, but the gunplay is always tight and satisfying, especially when you experiment with new weapons.

It’s amazing how many different things can happen to the human body depending on what you fire through it. My favourite environment was unquestionably the open space areas, where you’re forced to fly several hundred meters through the cold, silent depths of nothingness. No other game has captured the terrifying loneliness of an EVA like this.

The other huge improvement to the game is the new weapon and inventory system, though it comes at a cost. There are literally hundreds of different weapon types to be crafted at the benches, and you’ll experiment with new varieties every ten minutes or so, as the benches are liberally spread throughout the levels. Each weapon and attachment has a wildly different effect to the rest, and it’s obvious the game designers are fans of Insomniac’s inspirational armouries. Unfortunately I discovered an absolutely killer weapon combo about half way through, that made even the toughest of encounters a bit too easy — and this was while playing on the hardest difficulty level, ‘Impossible’.

As far as PC-specific features go, there’s a decent selection of graphics options. Anti-aliasing in particular has an impressive range of choices including SMAA and FXAA, but MSAA it out thanks to the deferred rendering engine. It ran beautifully on max settings at 1080p on my 4.5GHz CPU with dual GTX 670s (as it freaking well should) and frequently pleased my ocular system with breathtaking vistas, despite the occasional low res texture. The gorgeous visuals are exceeded by the game’s award-winning audio, with horrifying ambient effects and a soundtrack that perfectly matches the tempo of the onscreen action. This is the kind of OST worth buying on iTunes.

Based on my experience with the series, the Dead Space 3 that I played perfectly fitted my expectations for the game, but it could be very different from the Dead Space 3 that you end up playing. Played solo, the continued reliance upon the monster closets and flickering lights to raise your pulse will probably become a tedious grind of familiarity, rather than the tense moments of co-operative pants-soiling that I experienced. Fans of the series should still check it out just to see where the Black Markers take Isaac this time around, but to see the game truly shine, they should bring a friend along for the ride.

Good:

• Visceral are the masters of creating horrifying spaces that reek of authenticity, and Dead Space 3 is packed full of them.
• Co-op makes all of the old tricks fun again.
• A stunning audio-visual experience.
• The campaign delivers around 20 hours of value.
• Weapon customisation is actually fun as a result of the wildly different effects, rather than the meaningless grind found in other games.

Bad:

• The old tricks mightn’t be so much fun when played solo.
• Some might not enjoy the trend towards action in the final third act.
• Some weapons feel very overpowered, especially when you and a pal are both packing serious heat.
• No multiplayer means that once the campaign is done, there’s little reason to return.
• The much vaunted player-specific hallucinations don’t crop up enough to warrant the attention they received.

Full disclosure: Bennett is currently shacked up with an EA employee, but any concerns of bias can probably be dismissed by his glowing 4/10 review of the recent Medal of Honor over at Ausgamers.

At a recent EA event, games.on.net had the chance to sit down with Visceral’s Dave Woldman, senior producer on Dead Space 3. Woldman discussed Visceral’s approach to horror, how they’re planning to keep the game scary, and how truly successful horror is about the absence of combat, and more about what you don’t show rather than what you do.

Woldman also touched on the Dead Space community’s reaction to the new game, saying “People are resistant to change. They don’t know about the thought processes we go through, and that what we’re trying to do is evolve the franchise, deliver more of what they want in new and creative ways.

“I think the natural resistance to change is something expect, especially when you have such wonderful, rabid, passionate fans that we have.”

You can watch the video below, or download it in HD from our file library. A YouTube mirror is also available.

Dead Space 3 launches in Australia very soon — February 7, to be exact — and it’s bringing with it a whole new approach to co-op, where the game dynamically responds to the presence of another player by adjusting the storyline and levels to suit. games.on.net recently had the chance to sit down with producer Shereif Fattouh to discuss how they’re going to keep the game scary even with another player around to comfort you.

EA provided games.on.net with flights and accommodation in order to attend the Dead Space 3 preview event in San Francisco.

At the recent EA summer showcase, using the aid of his nanosuit’s cloaking mode, Bennett Ring snuck into the dressing room of Michael Elliot Read, Producer on Crysis 3. Unfortunately the cloak energy ran out just as Michael was taking his pants off, so to defuse this awkward moment Bennett quick-fired off a list of things he’s been wondering about EA’s upcoming sci-fi stealth blaster. The result was this interesting insight into Crysis 3’s approach to linearity, local matchmaking, hardware requirements and more, with not a peeping tom violation in sight.

GON: Crysis 2 played best when you were in the open areas, but there were also frequent linear sections. How are you approaching that mix of linear versus open in Crysis 3?

Michael: A lot of people came back and said Crysis 1 was very open, in fact a lot of people called it open world, which is completely incorrect. When you look at Crysis 1 and the areas you had to deal with, you’d come up on a vista and you’d be able to see a beach, then a village with trees, and mountains in the background, creating an open visual reference to where you could go. In Crysis 2 there were other areas where, yeah, it did open up a little bit, but dealing with the urban grid and the skyscrapers of New York City created a closed feeling to where things were, and where you could and could not go.

In Crysis 3, we’ve basically taken New York City, taken jungle elements, and thrown those two into a pot and combined them together. You have these large vistas, you have these large open spaces, but still you’re able to traverse through these. We’ve opened that up with these having these more open spaces, for exploration and secondary objectives, which encourage people to explore the maps.

GON: Variety of enemy type and AI is crucial in shooters these days, with games like Borderlands 2 having dozens. Crysis 2 didn’t have a wide range though — how are you improving it this time around?

Michael: There’s been a huge focus with AI and how we handle that. Back in Crysis 1, the AI had multiple alertness levels, from green to yellow to red. In Crysis 2, they only had two states – green and then straight to red. In Crysis 3 we’ve added the middle mode back again. On top of that we have some new tools in the engine that are available to us, like the AI navigation mesh. That allows our level designers to build out maps, and the system then says the AI can walk in these specific areas. You’ll find when you’re playing that when you come into a map, you can take 50 different people playing through it, and each experience will be very different depending on how the AI reacts and where it goes. We’ve done a host of tweaks to the AI and it’s really coming into its own now. It’s feeling really good.

GON: The multiplayer of Crysis 2 had some interesting elements, but didn’t really take off with players. What are you doing this time around to make Crysis 3 stick?

Michael: We’ve looked at what didn’t work in Crysis 2. We’ve brought in large numbers of groups with the help of EA and their internal teams, and also brought in groups which we call “The Experts Club”. These are comprised of people who play on consoles and PCs, and who have and have not played other Crysis games. We brought them in and did a whole series of feedback sessions with them. We found out what would make the game easier for people to get into, what would increase the pace of the game. We did a number of things based on this.

One is taking the nanosuit and changing how the energy dynamics work, breaking those down into armour and cloak, having completely independent bars. That’s been a huge hit with people. We did a closed alpha back in October, and people saw the pace of the game increase. We’ve also added in things like auto armour modules, which come in after a few levels in the game. These can be used in a perk slot, which will automatically flick their armour on. It’s kind of like training wheels. But it takes up a slot, and as you unlock more of these perks, you can start choosing which ones to use. It’s about getting used to those aspects of the game, and that’s what we had found was a barrier to entry. The maps themselves were also tighter, now they’re significantly larger than Crysis 2.

GON: Your Hunter mode seems to be inspired by a mod called The Hunted. Are you familiar with this, and was it an inspiration for your game?

Michael: I am not familiar with The Hunted, that’s probably something for my game designers and where they came up with the mode. I know it’s something that had been toyed with for quite some time. As you might know, Crytek UK is the former Free Radical studios, which is known for multiplayer titles like TimeSplitters. These were things that they had talked about for a long time and really wanted to get into the game. Eventually they started pushing that forward. He doesn’t like me using terms like Zombie mode or Infected, but there is inspiration from a lot of different things. The Hunter mode really worked well with the nanosuit, and having the bow there as well, just seemed to work.

GON: How do you balance the use of the bow while cloaked in Hunter mode, to stop it being hugely overpowered?

Michael: There’s a few things that tie into that. One is that you’re not fully cloaked, you’re slightly visible. There’s lighting and stuff that we’ve had to tweak for each map so the suit isn’t the most predominant thing, but is hanging back on the edge of your vision. Even walking through water will show where it is. In addition to that, all the Cell troopers have the ability to know when a hunter is near. There’s a clicking sound that gets louder and louder as the hunters get closer to you, so you know there’s a hunter near you and it puts you into a bit of a panic. It’s those kinds of things. We’ve seen games go in a number of ways, from Cell surviving for the entire game through to hunters dominating. It throws it both ways.

GON: When Crysis 2 launched, Aussies had massive problems with the matchmaking. Has this been a focus for Crysis 3?

Michael: Yes, there’s been a lot of stuff done on the back end and the way that we handle servers and technology, which has been influenced by EA’s experience as well. With Crysis 2, yes, there were a lot of complaints around the server technology and people having connection issues, not just in this area of the world, but pretty much all over the world, which we were never able to fully go back and resolve. Now we have the blade server technology which a number of EA titles are using, and PC gamers will have dedicated servers, which means servers anywhere in the world. There’s matchmaking on the console, I can’t really speak specifically on how that’s going to work, but I’m pretty sure that’s going to be a much more refined experience.

GON: People still make jokes about whether their PCs can run Crysis 1, but Crysis 2’s graphics options were basically non-existent. What are the plans for the level of configurability for Crysis 3?

Michael: There is a lot of configurability in Crysis 3. Let’s rewind a little bit, we basically had to take the engine, break it down and allow it to work on consoles and PCs. That was our first foray into the console world. We managed to do that with Crysis 2, but unfortunately there were some things that ended up getting left behind. One was the day-one high-res texture pack and DX11 support, both things that we really wanted to have there on day-one. Those are going to be there day-one for Crysis 3. In terms of options, we have full AA modes, sliders for pretty much everything, they’ll all be there in Crysis 3.

GON: Will it have a benchmark mode?

Michael: I can’t comment on that. Possibly, I’d have to look into that one specifically, but nothing I can comment on right now.

GON: We played the recent multiplayer alpha, and it was very demanding on even beastly PCs even on medium detail. Why was it so demanding? Is that going to be an indication of the hardware requirements we can expect of the full game?

Michael: The performance of the alpha, we were in an alpha stage where it hadn’t been optimized for performance. That was more of a technical test more than anything, to prove that our server back end was working and that we were able to support this properly. Also to get feedback from the client as well. Yes, there is definitely going to be a step up in hardware requirements, we’ve released the minimum specs and recommend specs. They’re relatively high. That’s what we’re moving forward with.

GON: Do you think aiming for relatively high PCs will be a benefit? We see so many games targeting low end systems and they miss some of the huge benefits of newer PC hardware.

Michael: If you look back on Crysis 2, people were like “this didn’t push my PC hard enough, and I didn’t have the options to do so”. We have a number of facets inside Crytek that we’re dealing with. One is the CryENGINE and things we need to do there. Not only pushing our own titles, but pushing support for our partners, and really finding out what we’re able to do and how hard we’re able to push that. Just because consoles are there and we have some eight year old technology doesn’t mean we should be holding back. We’ve really looked at that, and asked how hard can we push this engine, and I think gamers want a reason to upgrade their hardware. Crytek’s been known for several years to do that. Crysis 2 was a big challenge for us, having to pull back for consoles, to allow our engine to be a cross-platform engine, but now we’re not being held back by that.

Click here for our recent hands-on with Crysis 3.

As Crytek’s first multi-platform game, Crysis 2 on the PC wasn’t all that it could have been. I’m not just talking about graphics options or eye candy – even the gameplay itself had many of its finer nuances tuned out to create a more accessible experience for the console crowd. Enemy AI was only capable of two alert states versus the three in the original, and most areas of the game had a singular path between beginning and end.

While it proved to be a hit with controller crowd, PC gamers weren’t too impressed. During my recent hands-on with Crysis 3, the developer went to great lengths to emphasise that Crytek had learned its lesson, and that yes, Crysis 3 will once again feel like a true PC experience. I spent 30 rather enjoyable minutes discovering whether or not these claims were true.

The demo commences at the bottom of a set of stairs, with a couple of ammo boxes ripe for plunder. The only way out is up, leading to a large hallway that appears to be an abandoned train station. Rusted carriages line the tracks, while enemy soldiers man elevated walkways. One unlucky soul is positioned with his back to me as I enter the area; ramming my knife through his throat silently removes him from the complex equation of death that I’m about to solve.

Using my suit’s visor, I scan the area, zooming to take in the length of the long hall. Enemies in my line of sight are first identified with a highlighted outline, and the longer I hold my scanner on them the more info appears about them. I can see immediately that several of the guards are heavily armoured, while two automated turrets further down the room are just begging to be hacked, turning on their unsuspecting masters.

My first attempt of the room uses the successful “shoot anything that moves, grenade it if it’s still moving” approach that I employed so successfully in the last game. Within seconds of being spotted, the enemy AI has called out to his pals, sending them all in my general direction. They’re a talkative bunch, telling me where they think I am, and how they’re going to get to me. As three of them converge on my crouched hiding spot behind a barrel, I activate my suit’s armour mode, managing to pick one off before the other two send me hurtling back to the reload screen.

Climbing the stairs once again, I decide that stealth is the better option. Again, Mr Back-Towards-Me swallows my blade with the nape of his neck, but this time I head downstairs. Activating the cloak mode with its new unique button (major powers now each have a unique key press, making them easier to activate) I immediately notice how much longer the cloak ability now lasts.

Not long enough though; as I cross half way across the room I wink back into visibility, immediately spotted by one of the walkway guards. Rushing to more cover, I re-engage cloak and move on, watching as the enemy AI converge on my last known position. Soldiers fast-rope through the ceiling, providing me with the perfect opportunity to test out the air-burst bow ammunition, mowing three down with just two shots, before I’m picked off yet again.

I eventually clear the room after five tries, and each time the AI reacts to my actions totally differently. I end up completing it by playing much more cautiously than I’ve ever had to play a Crysis game, approaching it in the same way I did Dishonored, with stealth. I silently pick off enemy soldiers one at a time, keeping close track of their movements, until I’ve whittled them down to a small enough group that I can go in all guns blazing. Despite getting my butt resoundingly handed to me by the AI, it rarely ever feels like it’s cheating; each time I die I know it’s because I’ve made a mistake, not some super-human AI ability. Eventually I make it to the next major area of the demo with a sigh of relief, and encounter a huge outdoor area with head-high grass.

As the grass bends and moves with the wind, it’s easy to appreciate the leap in graphical fidelity compared to its predecessor. As I sit there, enjoying the detailed view of the vast space before me, with god rays lighting up the decrepit ruins around me, an alien creature uses the high grass to sneak up on me and tear me to shreds. The next fifteen minutes are some of the most tense first person shooting I’ve experienced in years, as I navigate through lengthy patches of this grass, with only the shrieking of incoming aliens to warn me of my impending doom.

By the end of the demo I’ve also expressed my inner nerd rage via a satisfying turret section, and taken in a spectacular train-ride through a heavily scripted sequence. I’ve also had more fun playing this demo than any singleplayer first person shooter of 2012 – that’s saying a lot considering the demo took place in a crowded, noisy, gaming event. If my short hands-on is anything to go by, Crysis 3 has abandoned the tired linear shooting gallery formula that we’re all so sick of, instead presenting gamers with large open-ended arenas chock full of clever AI, leaving it to us to figure out how to get through.

It might have taken three attempts to get it right, but Crysis 3 could be the one that finally delivers on the series’ promise.

Attending the recent EA Summer showcase, there was a recurring theme amongst all of the gaming press; everybody loved playing SimCity. And I mean everybody. Guys who prefer wasting zombie Nazis in BLOPS2 were having a ball managing the waste system of their little toy city; people who enjoy living the life of a drug trafficker in GTA were instead figuring out how to handle the traffic jams of their tiny town.

I’m more of a Town’s secretary than Mayor – I played way too much Settlers 7, and something like SimCity seemed way out of my pathetic micro-management league. Yet after completing the introduction mission, I had mastered SimCity’s intricate and layered ecosystem. That’s the real beauty of the new SimCity – despite being an onion of complexity, with layers upon layers of stats, the Glassbox engine includes visual indicators that make understanding what’s going possible in a single glance.

Before you even lay down the first slab of concrete, it’s wise to figure out where the natural resources are. Doing so is a matter of simply clicking the right menu item and the landscape dissolves into a transparent box, revealing the underlying coal seams and gas ponds below, just waiting to be extracted by your eager minions. It’s clear, and easy to understand – almost like an infographic view of your resources.

Once you’ve settled on the ideal place to destroy nature, it’s time to roll out roads with the new road laying tools. These have been totally revamped, making it much simpler to create gentle curves or sweeping bridges. The refined tools end up giving more intricate road networks when you’re done, but are simpler to use in the first instance. Try going up a cliff and the tool won’t let you – all roads are limited to the same gradient as the real world, with a maximum of around 35 degrees. With a few major roads splitting off into dual lane streets, I’d soon laid the foundations for my very own mini Big Apple, using a grid system to keep my OCD impulses happy.

Time to lay some residential areas then – once again, the tool automatically shows me where I can place them, and it’s just a matter of clicking and dragging to fill the spaces. And then I just sat there for several minutes, watching the dozens of workers driving up to the empty lot and begin building. It’s astonishing how detailed the game is, yet it doesn’t bring the demo PCs to a screeching halt. A few commercial and industrial centres later, and I’ve got a living, breathing city right before my eyes. It’s filled with inhabitants, and clicking on each one tells me where they’ve been and just how content they are with my creation.

As my city grows, problems like crime and pollution start to swell, but again I can click a single button to reveal an infographic, transparent version of the city. Brightly coloured areas show me exactly where I need to fix problems; the sewerage view shows brown blobs surging below each street, forming a block around the sewerage plant and indicating that I need to expand my poop-carrying capacity. Red buildings show where criminals have set up shop, though I should have noticed the environment filling with graffiti before now.

It’s this ability to strip your creation back to a single layer, exposing only the information that is most relevant, that will make managing cities in SimCity even easier than ever. The fact that your city truly feels alive, with realistic animations for any action you can think of – from fireman spraying out fires to garbage dumps burning off waste – means watching the city develop never gets old.

If the idea of building a city is about as appealing to you as reading a street directory, SimCity could be the game that changes your mind. The clever tools unleash more possibilities than ever before, but their focus on usability means it should never become overwhelming without a 400 page manual nearby. If it can convince a Battlefield 3-loving, Hitman Absolution slaying player like me to buy the full game, chances are there will be a whole influx of new players come SimCity’s March release date.

BenQ’s XL2420T monitor set a new benchmark for 24-inch gaming monitors, not least in the pricing department. At $500 it’s around twice the price of competing displays, but your wallet won’t feel molested when you wrap your loving gaze around its compelling image quality, with razor sharp response times and image presets custom built for first person shooters.

Sadly half a grand is still out of the reach of most mere mortals, despite the amazing feature set, so BenQ has released another 24-incher that won’t see you sleeping in the dog house if your other half finds the receipt. The RL2450H retails for just $199, yet BenQ claims it shares many of the same qualities that its pro-gamer cousin includes. Let’s see if it can live up to the family name.

Specs:

• Screen Size: 24”W
• Aspect Ratio: 16:9
• Resolution (max.): 1920×1080
• Pixel Pitch (mm): 0.276
• Brightness ( typ.): 250 cd/m2
• Contrast ( typ. ): 1000:1
• DCR (Dynamic Contrast Ratio) (typ.): 12M:1
• Panel Type: TN
• Viewing Angle (L/R;U/D) (CR>=10): 170/160
• Response Time(Tr+Tf) typ.: 5ms, 2ms GTG
• Display Colors: 16.7million
• Color Gamut: 72%

The ‘RTS Monitor’

BenQ enlisted the aid of StarTale, a bunch of Korean gamers famous for dominating the StarCraft II scene, to help make this the perfect monitor for RTS gamers. Now, given the slightly slower pace of RTS games, you’re probably wondering what makes for a good strategy monitor. The answer is a dynamic contrast booster which BenQ calls the “Black eQualizer”, which is meant to allow you to spot bad guys tucked away in the shadows. It basically adjusts black levels on the fly, which gives it a slightly washed out look — however, it also actually works as advertised.

Most pro-gamers already run with jacked up contrast or brightness to achieve the same effect, but on the RL2450H you can achieve the same result without making your game look quite anywhere near as bad.

Another nice touch is the ability to scale the screen to various resolutions other than the native resolution of 1920 x 1080. If you find your old GTX 560 is struggling with Far Cry 3 at 1080p, simply bump the resolution down to 1650 x 1050 and the monitor will only display those pixels, rather than stretching the resolution across the full 24 inches. It’s great for those who don’t have the grunt to run at HD all the time, yet hate the blocky image that usually happens when you run your LCD at a non-native resolution.

Frames and construction

At this price we didn’t expect the fanciest of frames, and weren’t surprised when we discovered a rather rudimentary stand. There’s no way to adjust the height of the display, which could prove to be a neck-straining bother for those with gaming dens that ignore ergonomics. It can be tilted though, which will be handy for those with a monitor stand just a tad too high.

Inputs are taken care of by a single HDMI, DVI-D and D-sub input, so you won’t be able to attach both your PC and consoles without using a HDMI switcher. Again, at this price we can’t expect more than the basics. The On Screen Display (OSD) was easy enough to navigate, but only includes the most rudimentary of tweaks; professional designers won’t find the colour calibration options up to snuff for their demands. Physical buttons are used to move around the menus, with touch sensors prohibitively expensive at this bargain basement pricing.

BenQ has fitted the RL2450H with the older style of TN panel rather than the newer, more expensive IPS panels that are now becoming the norm. As a result the viewing angle isn’t nearly as good as other IPS 24 inchers, but this issue is moot for gamers, who wouldn’t share their gaming screen even with a Glock pointed at their noggin.  The benefit of the TN panel is a rapid-fire response time of just 2ms, which leaves most IPS screens for dead.

We noticed absolutely zero ghosting on the RL2450H, with crisp motion even when playing hyperactive games like Quake 3. Unlike its more expensive brethren, the TN panel used here is limited to 60Hz rather than 120Hz, which means it can’t do 3D. While 120Hz is incredibly smooth, the vast majority of gamers don’t have the horsepower to churn out such a high frame rate, so again this problem won’t be noticed by the majority of the population. On the subject of 120Hz, we suggest avoiding seeing games run at this speed, as it’ll make your lowly 60Hz screen seem positively sluggish. And once you’ve seen the incredible fluidity of 120Hz, it’s impossible to un-see it.

Colours

Another issue with TN panels is their colour reproduction, and the RL2450H scrapes by with relatively average colour quality. Colours appeared to be overly saturated, leading to a very garish image quality on the default values. These can be toned down a little by tweaking some of the image options, but overall we’d describe the colour reproduction as just a little too vibrant. Having said that, we prefer a monitor with richer tones than those with flat, dull colours, so it’s not a deal breaker of an issue.

The LED backlighting looks fine during games, but running a pure black screen did indicate a few uneven areas that were a little too bright. However, grey scale test results were excellent, showing that this monitor is a beast when it comes to contrast performance. It’s not too surprising given the fact that BenQ is pushing contrast as its defining feature, but it’s always nice to see the real world performance living up to the marketing claims.

At this price point we didn’t expect retina-gouging performance, but BenQ has delivered yet another able gamer display despite the meagre budget. Compared to similarly priced offerings this screen offers superior contrast performance, a crucial feature for making out detail in darker games. The RTS mode is actually useful for competitive play, and not just a load of marketing bollocks, but the RL2450H is just as comfortable playing shooters and action games as it is strategy games. Colour quality might not be perfect, but with a little tweaking it’s more than adequate for all but the most demanding of users.

It’s still no match for the XL2420T, but at less than half the price it’s also a much easier purchase to justify to the other half. Recommended for those on a budget.

Good:

• Won’t cause your wallet to implode
• Excellent contrast performance
• Impeccable response time with no ghosting
• Solid overall image quality

Bad:

• Average colour reproduction
• No height adjustment
• RTS mode is a little washed out for our liking

It might have been plagued with problems, but Hitman: Blood Money still has a cult following to this day. Each mission took place in a huge, open killing ground, where the method of death was limited only by the player’s twisted imagination. The game provided the tools and the environment, but it was up to you to decide how to manipulate them through hours of observation and experimentation, and it’s still revered as one of the finest examples of sandbox gameplay. Powered by the meagre muscle of the PlayStation 2 and Xbox (with the 360 version really just an afterthought), it’s no wonder virtual killers were fingering their garrottes with glee when Hitman: Absolution was announced. Today’s consoles are exponentially more powerful than their predecessors, so we could only imagine how much bigger and more densely packed with death traps the new game would be.

Instead IO Interactive used the power of the PS3 and 360 to deliver the most claustrophobic Hitman game yet.

Gone are the huge playgrounds of Blood Money, with each level now broken down into bite-sized murder-chunks. The reasoning behind this compartmentalisation is obvious as soon as you lay your eyes on Absolution’s luscious levels. By making each area much smaller, the game can cram more detail into the environment and the objects that fill it, without causing today’s consoles to burst into flames. The PC version is simply spectacular, using hardware tessellation to deliver a level of detail over and above the consoles, and the crowd technology is still as impressive as it was in the last game, with hundreds of people filling clubs and marketplaces.

The NPCs have been brought to life, far more believable than the robotic automatons of the past, spewing out lengthy reams of voiced dialogue provided you stick around long enough to listen in. In Blood Money, it often felt like you were a kid with a magnifying glass looking in on a huge robotic ant farm, but Absolution’s areas feel much more alive and real, albeit at the cost of scale and openness. While the look of the game is mostly breathtaking, IO has gone a bit crazy with the lighting and bloom stick; even Agent 47’s bald dome casts off enough glare to distract overhead pilots. The eye candy also comes at a cost, occasionally bringing my i7 3770K (overclocked to 4.5GHz) with dual GTX 670s to its knees, often in areas that should run beautifully, such as tunnels.

Due to the smaller size of the areas, many levels follow a single linear route, making it less about experimenting with different pathways and more about following the obvious cues to the next checkpoint. Every now and then it’ll open up into a larger area like a marketplace or apartment building, where hints of the old Hitman experience shine through, but even these areas pale in comparison to the hectares of space virtual hitmen are accustomed to exploring.

Making it even less about player choice is the fact that you don’t get to choose which weapons to start the mission with. Instead you spawn with whatever the level designer has deemed suitable — but you don’t have to use them. There are usually three or four traps squirrelled away in the environment, and they’re not always easy to find; in one section it took me an hour of reloading before I found a certain poisonous fish that I could use to poison my target.

This has a huge impact on the playstyle — whereas in the past you could choose which weapons to take, and the levels were littered with various options of despatching your target, now it plays out like an easter egg hunt. Walk through the level, stumble upon the obvious trap, then trigger it when the target is near. The game doesn’t even let you use weapons that you ended the last mission with — if you left the last mission bristling with items tailor made for mass murder, they’re all gone by the time you start the new mission, even when both missions are directly connected in the game’s timeline. This lack of continuity is strange problem for a game that prides itself on bringing a strong narrative to the series.

The choice to use pre-rendered cut-scenes means that other continuity issues constantly rear their ugly head. Countless times I opened a door in-game, garbed in a stolen police officer’s work blues carrying a police revolver, only for the game to jump to a cut-scene of me entering the room in Agent 47’s trademark suit armed with a silenced automatic pistol. It’s jarring and it’s basic film school stuff, so it’s baffling that a game that places such a heavy emphasis on narrative ignores one of the basic tenets of the storytelling process. If you can ignore the issue, the storyline is actually rather decent. It’s not going to win any Emmy’s, but it’s also not cringe-worthy. We’ve all known that – deep underneath the coat of blood of his recently eviscerated victims – Agent 47 has a good heart, and Absolution explores the theme of how many lives he will take to save a single life.

Chances are you’ll take a lot more lives than you want, as playing the game in stealth mode is brutally difficult. Demons Souls is an utter cake walk compared to stealthing through Absolution on the higher difficulty levels, as the linear level design and aggressive AI makes it incredibly hard to get through most areas. The new disguise system makes stealing uniforms basically worthless, as enemies garbed in the same attire will spot you in seconds unless you use your Instinct power to hide your face, but it burns out in just a couple of seconds. Throw in checkpoints that are harder to find than a working follicle on 47’s head, and this is one tough sucker.

Like Demons Souls, the sweet, sweet sensation of satisfaction of getting through a level makes it all worth it, but you’re going to need the patience of a saint to ghost your way through. Blasting enemies is infinitely simpler, and the improved controls and weapon feedback makes this a very serviceable third person shooter.

This gunplay can be used in the online Contracts mode, where you must kill a target in an area according to a plan built by other players, but most players will probably set up tricky silent assassinations instead. It’s an interesting mode that will likely die out rather quickly, as the joy of setting up or achieving a kill is still limited by the linear levels.

If it sounds like I’m disappointed in the design decision to lessen the scale of the game, it’s because I am — but that doesn’t necessarily make Absolution a bad game; it’s just very different to past Hitman experiences. Experimentation is now replaced by observation and persistence, and it feels much more like a Splinter Cell or Metal Gear game. Yet there’s still the feeling of sweet satisfaction after clearing a level without killing anybody, and in some ways this is even more pronounced thanks to the fiendishly clever AI and devilish difficulty, provided you’ve got the tenacity to retry levels dozens of times.

Many gamers will also appreciate the increased attention to detail, and won’t miss the head-scratching problem solving required in the past. If only IO Interactive could have accomplished both, combining the open nature of the previous game with the lush minutia of Absolution, it’d be the perfect Hitman game. Instead we’re left with a much more focused assassination experience, which sacrifices emergent gameplay for blockbuster presentation.

Good:

• A visually gobsmacking game which makes the most of the PC’s powerful graphics capabilities
• AI is much more believable thanks to better behaviour and a wealth of dialogue
• Pulling off environmental kills is still incredibly satisfying
• Gunplay is now a viable – and enjoyable – way to clear the game

Bad

• Doesn’t feel like a Hitman game
• A narrower experience than the sandbox gameplay of its predecessors
• Incredibly hard to successfully stealth your way through on higher difficulties
• Very demanding hardware requirements if you want all the eye candy

Using your mobile at a gas station can cause the pump to explode into flames. Increasing the amount of memory on a video card from 2GB to 4GB will increase game performance. Blue whales are the largest living things on Earth. Which of these three facts is actually true? The answer will probably surprise you – they’re all bollocks. Mobiles have never caused a fire at a petrol station, the largest living thing on the planet is technically a mushroom colony in Oregon’s Malheur National Forest, and doubling your VRAM from 2GB to 4GB has absolutely no impact on performance… for 99.99% of users.

The vast majority of today’s video cards have enough memory that size isn’t an issue, with mid-range cards equipped with 1024MB while premium jobbies are decked out with 2048MB. Sure, running at 16x antialiasing with 2650 x 1600 resolution will hit the 1024MB limit of a GTX 650 Ti, but the GPU doesn’t have the horsepower to process that kind of quality anyway. Cards that do have the grunt to churn out such cracking graphics, like the GTX 680, have the full 2048MB of memory, and that amount is plenty even when running at such stupidly high settings. Sadly most gamers don’t realise this; they simply see a big number on the box and thinks it means they’ll get a squillion frames per second in Battlefield 3.

The more-memory-myth

This myth is perpetuated by AMD in particular, endowing its products with copious amounts of memory well over the upper end used by games of the period. Currently AMD’s Radeon HD 7970 ships with a whopping 3GB of GDDR5, yet PC developers aim for a video memory amount of between 1024MB and 2048MB, depending on the card. I can only conclude AMD does so because gamers believe it makes for a better experience, resulting in more sales.

This same company often trots out the myth that folks packing triple monitors need the extra video memory due to the higher demands of running at 5760 x 1080. In theory it makes sense – the larger the image, the more memory needed for textures. Antialiasing also gulps up memory like a Tour De France rider in a steroid factory. However, in reality the memory bandwidth – that is, the speed at which the memory can be accessed – is much more important than how much of it there is. Less memory running at faster speeds will generally beat more memory running at lower speeds. There’s a reason that a 2GB GTX 680 generally beats the 3GB Radeon HD 7970, even while running triple screens.

Let’s put it to the test

But what about a video card like Gigabyte’s new GV-N680OC-4GD, which uses the powerful GTX 680 GPU to drive a whopping 4096MB of memory? How does this compare to other GTX 680 cards packing 2GB, but otherwise identical? The memory bandwidth on both versions is identical, with an effective memory frequency of 6008MHz running over a 256-bit memory bus. Does doubling the memory have any impact on game performance?

We ran one of the most demanding games currently available to find out. Battlefield 3 is known for using as much memory as it can get its battle-fatigued mitts on, so we ran it at three resolutions including triple 1920 x 1080 screens, and also compared it with an NVIDIA GTX 680 reference card overclocked to the same speed as the Gigabyte card. An AMD Radeon 7970 was thrown in to see whether its 3GB of memory would help.

As you can see, the extra memory made absolutely no difference. In fact, the 2GB version was 1 frame per second faster at 1080p, but this is within the limits of variance. Buying a 4GB card to play today’s games just doesn’t make sense.

We ran one more test, just to double check the results, and this time left the Radeon HD 7970 out so you could focus on 2GB versus 4GB. Dirt 3 uses DirectX 11 effects to create one of the best looking racers around, and serious drivers often race with three monitors to get that “in-cockpit” sensation. Will a 4GB card help them out?

Once again it’s plain to see that doubling the VRAM has had no impact at all.

And yet, we’re about to toss everything we just said out of the window. Buying a video card with more memory can be a good idea, but only if you’re buying with an eye on future gaming performance. It’s inevitable that in the next few years 2048MB will become the bare minimum acceptable for gaming, with the preferred amount leaping to 3072MB. If you’re buying your card today with the idea that it’ll last you another three or four years, it’s wise to buy a model with more memory to meet those future demands. However, if you’re like me and silly enough to upgrade every year or two, the added memory isn’t worth it unless it costs exactly the same as a card with less memory.

So the doubling of memory on the Gigabyte GV-N680OC-4GD doesn’t make much difference in today’s games, but what about the rest of the card? Being based on NVIDIA’s GTX 680 GPU we know that it’ll perform basically identically to every other GTX 680 on the market; this is the fastest single GPU product currently available. Gigabyte has given the card a very slight factory overclock, increasing the GPU core speed to 1071MHz, which is about the same as competing cards from Zotac and ASUS. Where it differs is the use of a triple fan Windforce cooler. It might not look like much, but those three fans are remarkably efficient at moving heat away from the GPU. We ran FurMark to test the fan noise under load, and were blown away at the 46dB rating, which makes this one of the quietest GTX 680′s on the market.

With street prices hovering around $650, the GV-N680OC-4GD costs around $100 more than Gigabyte’s 2GB GTX 680 products. Whether or not you want to pay the extra amount totally and utterly depends on how long you think you’ll be using the product. If you plan on upgrading to next year’s GTX 780 when it comes out, stick with the 2GB version and spend the $100 you’ll save on another SSD. However, if you’re holding off for several generations until the GTX 980 releases, the extra memory will come in handy. By 2015, when Battlefield 6 requires more video memory than that found in an entire 2012 PC, the 4GB of onboard memory will probably come in mighty handy.

Good:

• Great for future-proofing if you upgrade infrequently
• Overclocked to 1071MHz
• Whisper-quiet at 46dB, one of the quietest GTX 680′s on the market

Bad:

• Essentially zero performance improvement over stock 2GB version of the same card under current gaming conditions

Ever been frustrated by being shot as you run into cover in Battlefield 3? Ever wondered why you can’t look down the scope while crawling in prone mode? What about why games like Battlefield and Medal of Honor don’t feature locational VOIP, or what system requirements you’ll need to run Warfighter on Ultra?

We’ve been speaking to Kristoffer “Hoffe” Bergqvist, Creative Director of Multiplayer from Danger Close Games, and ex-DICE employee of six years on the Battlefield series, and he’s given us all the answers. Read on for our massive interview on all the nitty-gritty details on Medal of Honor: Warfighter.

games.on.net: The last Medal of Honor game’s multiplayer was created by DICE. Given that DICE has a lot of experience with multiplayer, why did you move the development of Warfighter’s multiplayer back in-house?

Kristoffer: We wanted it to be a cohesive experience, for singleplayer and multiplayer to share features. But we also wanted to share the unique tone that is in Medal of Honor. For that to happen we needed to be in the same studio. It was a great opportunity – there are still people in the Medal of Honor team who worked on the first Medal of Honor game with Spielberg fifteen years ago.

games.on.net: We did notice a disconnect between single and multiplayer in the last game. Do you think you’ve solved that this time around?

Kristoffer: Yes. Control input is the same, with peek and lean in both single and multiplayer. It’s also about the tone and style of the game. If you learn to handle the rifle in singleplayer it will handle the same way in multiplayer, and the transition between the two halves is much easier.

games.on.net: On the subject of player movement, during our demo we noticed that players can’t look down the scope while crawling in the prone mode. Can you talk about the reason behind this design decision?

Kristoffer: When you’re in prone, we wanted you to make a conscious choice before you go into prone. Do I want to be mobile, or do I want to be stealthy? It all came from there – we figured that if you want to move in prone you sacrifice the ability to look down the sight. It also makes no sense from an authenticity perspective. You can’t look down a sight while you’re crawling in real life.

games.on.net: We also noticed bullet drop is less pronounced than Battlefield 3. Can you talk about the reason behind this? Is it based on real world ballistics?

Kristoffer: We start with real world ballistics. Rather we start with the entire experience of the gun. What does it feel like to shoot – we want that to be as realistic as possible. We’re working with special operators, a bunch of them are snipers, and we’re trying to make the sniper experience as real as possible. We don’t have the combat distances that Battlefield has, so bullet drop is a lot less pronounced.

games.on.net: Is the bullet drop technically accurate, or is it exaggerated compared to real life?

Kristoffer: We start off as technically accurate. But it’s a game, it has to be fun, it has to be balanced.

games.on.net: What were the major flaws that you identified from the last Medal of Honor game that you wanted to resolve this time around?

Kristoffer: I’ll answer that from a multiplayer perspective. One of the things we knew from the start was that we wanted to be one, cohesive experience, which we’ve already talked about, but that was a big thing.  We also wanted to expand on the depth of the game, to be a game that you played for months and months and years. We did that in a couple of ways. Firstly we expanded the content a lot. Six multiplayer classes instead of three, classes all have unique support actions, they all have unique grenade types and weapon systems. Even things like run speeds. Just a lot of content inside the core gameplay.

We’re also adding endgame features, such as platoon versus platoon matches, which is our take on clan gaming, and trying to make that as accessible as possible. We also really want to support this game post-launch, making sure it stays up to date. We implemented systems for post-launch patching so that we can update stuff like weapon damage and recoil without deploying patches – we can do it server side. If we see that nobody is using a gun because it’s not powerful enough, we can bump it up without needing a patch.

games.on.net: On the subject of post-launch support, Frostbite 2.0 has excellent netcode that allows destruction and lots of players, but it’s also quite vulnerable to certain hacks as a result of the mixed client/server prediction. How do you plan on tackling those issues once the game has launched?

Kristoffer: We’re talking to DICE a lot. When they close a hole, we do it as well. We have removed parts of the code that were most exposed to hacking. It’s a very real problem, and we don’t tolerate it. We’ll do everything we can to stop it.

games.on.net: Will you have a dedicated team of people combatting cheating after launch?

Kristoffer: Yes we will. I can’t really talk about details yet, but absolutely.

games.on.net: And obviously you’ll have Punkbuster support for the PC version?

Kristoffer: Yes.

games.on.net: When the Unreal engine was first released and used by several games, they all tended to look quite similar. Frostbite 2.0 is the same tech that powers Battlefield 3. How do you make a game using the same engine, that doesn’t look just like Battlefield?

Kristoffer: It has been a key point for our development team since day one – how do we not look like Battlefield. How do we define Medal of Honor. I think every decision we made stems from there. It was important for our art director to come up with our own look.

games.on.net:  So what is the Medal of Honor tone? When you were setting out your design guidelines, what were some of the rules that you created that would remove your visual identity from Battlefield?

Kristoffer: There is a style that I can’t really put into words. Everything circles around authenticity, we’re looking at pictures that the special forces took when they were deployed overseas and trying to portray the world as they saw it when they were there, if that makes sense. I wish I had my art director here, as he could give much better answers.

games.on.net: EA’s made it really clear that Medal of Honor and Battlefield are intended to go head to head with Call of Duty. Of the two games Medal of Honor is probably more similar to Call of Duty. How do you appeal to that same audience, without coming off as a Call of Duty clone?

Kristoffer: We need to find our own route. We never looked at a competitor and said  we want to make what they’re doing, but better. We wanted to make our own game. We’re trying to bring in features that make our gameplay unique. The fireteam feature for example, which is kind of like taking co-op into multiplayer. It pairs you with a fighting buddy. This all started in discussions we had with the operators. We saw how well they work together in pairs, because they have one guy they always stick with for a very long time, and they get incredibly well synchronised. We wanted to include that.

So when you go into multiplayer, you’ll be assigned a fighting buddy, and you can always see where he is via a faint outline. It emulates the sixth sense these guys develop. You will see intel about what’s he’s doing, he’s hurt, he needs ammo. He also works as a spawn point – if you die and he stays out of combat you can spawn in on him. If he’s the more aggressive type, he can hunt down the guy that killed you, and you’ll spawn instantly. We’re all big fans of teamplay and want to reward that. We figure that with one team member to really rely on, it’s much easier to find that one person to play with.

games.on.net: So it’s just like the squads in Battlefield, but two men instead of four?

Kristoffer: Yes. What it allows us to do is share a lot more data between the two players. If you have four guys in a squad it gets harder to share that amount of data on that scale. We can also give you more direct feedback if you’re doing good teamplay-wise. We’re rewarding teamplay directly, so it’s easier for the player to see what they’re being rewarded for.

games.on.net: In terms of the actual gunplay – pulling the trigger, seeing the recoil, watching where the bullets land – how do you feel Warfighter differs from other military shooters, such as Call of Duty? Have you looked at their gunplay and decided what you wanted to do differently?

Kristoffer: We are doing a lot of things differently on the microscale. The things that are hard to talk about, that happen in the millisecond, like the tweaking of ADS speeds and headshot boosts. We wanted gameplay where shooter players felt like they understood the basics, we didn’t want to alienate anyone there. But after playing it, there’s something different. A lot of it comes from speeds, damage values, small tweaks.

games.on.net: Battlefield is known as having more recoil than Call of Duty – where would you guys sit between the two?

Kristoffer: We differ in the ways we handle recoil. There’s deviation, which is how the bullet doesn’t go where the crosshair points. Then there’s kick, which moves your entire gun. We rely more on the gun kicking. We have limited deviation, because it fits our combat distances and gameplay speed a lot better.

games.on.net: So does that make your first shot more accurate than in Battlefield, where you have to worry about deviation?

Kristoffer: Yes, the first shot is a lot more accurate. It’s going to be pretty apparent when you sit down with it and get a feel for it.

games.on.net: In terms of Frostbite 2.0, have you made any technical changes to the engine?

Kristoffer: We’re working a lot with hit detection.

games.on.net: Which a lot of people say is one of Battlefield’s flaws…

Kristoffer: We’ve been working on that. Our servers work differently because we don’t have vehicles so we can run faster, with more data and higher tick rates. We’re trying to create the balanced gameplay situation where you don’t feel like you get shot after running into cover. It’s a delicate balance but we spent a lot of time on that. We also worked a lot with minimising input lag. We put a lot of effort into this, especially for the console players. You’ll notice a big, big difference.

games.on.net: You’re locked to 30fps on console, right?

Kristoffer: Yes.

games.on.net: Is that the limiting factor when it comes to input lag on consoles?

Kristoffer: No, there are other things we can do. Like how we render our visuals. But we have shaved a couple of frames off the input latency. We bought this system that allows us to measure input lag very precisely – there’s this guy who builds controllers that allows us to measure it very precisely. So we hammered on that until it was good. We also did a lot more weapon customisation in the Frostbite engine, so you can swap out barrels and stocks. We also have the ability to draw the world, then render the character over the top of it in the UI. The end result is that in the respawn screen, we can show you the world, but we can also show you the character model. That was big for us, because we wanted to show each player the special forces player they’ve picked.

games.on.net: Any environmental updates to the engine?

Kristoffer: We worked a lot with water. You’ll see proper wave dynamics.

games.on.net: Battlefield 3 is arguably the best looking PC game at the moment. Do you think you’ve matched it in terms of visual quality?

Kristoffer: Yes, I’m really happy with the Warfighter looks. I look at it now and am just blown away. We worked closely with DICE with this, and I’m not the only ex-DICE guy in the office. We know each other pretty well, send resources back and forth, they help us, we helped them with Battlefield 3. We’re at this stage where so many games at EA are using Frostbite 2.0 so we can easily send resources back and forth. You saw the driving level in the Warfighter demo.

games.on.net: The level called Hot Pursuit? Why didn’t you just call it Most Wanted?

Kristoffer: (laughs) Obviously we couldn’t have done that without the driving talent we have at EA.

games.on.net: You’re matching in terms of visual fidelity, but what about system requirements? Battlefield 3 needs a beast of a PC to run at Ultra…

Kristoffer: It’s going to be about the same. If you can run Armored Kill you’ll be fine.

games.on.net: We looked at the graphics options in Warfighter’s demo today, and they’re identical to Battlefield 3. Are they final?

Kristoffer: I think so. They should be.

games.on.net: We were wondering if you might simplify them given that you’re going for a different audience.

Kristoffer: We are trying to streamline a lot of things. When it comes to PC graphics settings, I think most PC players want that fidelity in their options. I’m pretty sure what you saw is final.

games.on.net: VOIP in PC – is it in-game, or via an external system like Battlelog?

Kristoffer: It’s in-game. If you switch fireteams it’ll move you to that new group.

games.on.net: So you can only speak to your buddy?

Kristoffer: Default is your buddy, because we found that most of the VOIP traffic gets more relevant if it’s just the two of you. Then on certain screens like the end of round screen, we move everyone into the All channel, so you can gloat a little bit. We have plans to have server side options to allow the entire team to be on one channel in-game. I’m not sure that is done for launch though.

games.on.net: Will that be switchable by the user, so they can chat only to their buddy, or to the whole team?

Kristoffer: Yes, that’s the intent. But that’s all coming in now, so it might not be available for launch.

games.on.net: Any thoughts on having locational VOIP, so players can overhear enemies?

Kristoffer: They have it in Day Z, which I currently play quite a lot. It’s a lot of fun. I would love to experiment with that – we don’t have it in the game. We talked about since… we spoke about this when making Battlefield 2. I think it’s something that should be done pretty early in the development process. Do you remember Counter-Strike once tried it? The result was that players stopped talking to each other, so they reverted it really fast. You need to find a balance where that doesn’t happen. It’s a cool trade-off, should we talk or not, but you don’t want everyone to clam up and not talk, or go and use Ventrilo instead. So it’s a game design issue, not a technical issue.

games.on.net: In terms of the browser, will it be in-game?

Kristoffer: It’s in-game and via web browser, we have two versions.

games.on.net: Can it do region filtering? And why do two browsers? 

Kristoffer: It has the same features as the Battlefield 3 server browser, which has region filtering. I need to double check what it looks like in-game or in the browser. We wanted both because we feel that our player base is in two camps. A lot of players appreciate the convenience of having it through the web, but a lot of players like to do it from the game client. It’s one of the upsides of working closely with DICE, as they’d made all this before.

games.on.net: Is there anything else you’d like to add?  

Kristoffer: I really look forward to getting it out there and playing it with everyone.

Thanks to Kristoffer for chatting to us!

After our recent hands-on session with Need for Speed: Most Wanted, we managed to corner the game’s producer, Criterion’s Leanne Loombe, for a chat about why the PC version will make its console counterparts look positively ugly. Read on for all the specifics, including how they’ve leveraged DX11 for a 300% performance increase on PC, and exactly what sort of rig will be required to push it as hard as you can.

games.on.net: There’s a growing gap in performance between the PC and consoles as consoles get older. Does that create more work for you guys, because you need to spend more time on the PC version?

Leanne Loombe: Yeah, we have a dedicated pipeline for the PC version. The way that we have it set up means we can dedicate more time to PC. The PC version of Most Wanted is obviously 1080p, 60Hz, and we’re using DirectX 11 as well. That’s there for the PC gamers.

games.on.net: Do you know which features of DX11 you’ll be using?

Leanne Loombe: We’re primarily leveraging the increased efficiency of DX11 to give improved performance. The move to DX11 from DX9 has given us around a 300% improvement in rendering performance. This has enabled us to provide an improved image based lighting model over the console version, as well as allowing us to run with higher detail shadows and reflections. The increased power of DirectX 11 hardware has also allowed us to implement features such as real-time ambient occlusion, and light scattering algorithms which are absent in the console version.

games.on.net: What other changes have you made to the visuals in the PC version?

Leanne Loombe: PC supports a number of advanced graphical features including SSAO, light scattering, high dynamic range motion blur, high resolution textures, advanced specular lighting models, headlight shadow casting, enhanced VFX quality, and enhanced shadow quality levels. We also scale geometry detail and level of detail switching according to screen resolution.

games.on.net: How deep are the graphics configuration options on PC – are we talking a short list of three or four options, or a deeper screen with over a dozen? Can you tell us what options PC users will be able to tweak?

Leanne Loombe: Options are available to enable High Resolution textures, motion blur quality, shadow quality, SSAO level, reflection detail, visual effects quality, geometry detail levels, and light scattering.

games.on.net: Will Most Wanted need a very fast PC to run at the highest graphical details?

Leanne Loombe: A modern quad core PC with an AMD Radeon 6000 series or NVIDIA Geforce 500 series can run the game with highest details settings.

games.on.net: Will the PC version use peer to peer gameplay for the online mode, and what are the number of players it supports? Is it more players than console? Also, do you support LAN play?

Leanne Loombe: PC uses peer to peer and supports 12 players online, whilst console only supports 8 players. An internet connection is required for multiplayer, so LAN only play is not supported.

games.on.net: What controls will the PC version support? Can we expect full Xbox 360 controller support on the PC at launch?

Leanne Loombe: The Xbox 360 controller is fully supported and is the recommended controller for Most Wanted. We also fully support keyboard and the majority of DirectInput controllers, including steering wheels. Controller bindings are fully configurable in game.

games.on.net: Have you had to change the interface at all for the PC version?

Leanne Loombe: We have always strived to provide an accessible user experience, regardless of platform. We want the game to be fully playable whether you are hunched over a 15” laptop, or sitting on your sofa playing on your PC hooked up to a 50” plasma TV. As such minimal changes have been made to the PC interface, other than providing enhanced settings menus and a quit option. Origin In Game is also fully supported to allow you interact and manage your Most Wanted friends on PC.

games.on.net: Comparing the handling of Most Wanted to Hot Pursuit, are you aiming for the same feel?

Leanne Loombe: It’s brand new handling this time around. It’s unique for every single car, so every car has a different model. We wanted the player to feel like they were driving the real car because they’ll never get to drive those high-end sports cars in real life.

games.on.net: We just drove the Aston Martin in the demo and it felt quite heavy and sluggish. Is that because I was driving the Aston Martin, or is it part of the new handling, with cars having more weight?

Leanne Loombe: I’d say it’s probably because you were driving the Aston, where we tried to reflect the car’s weight.

games.on.net: How do you then balance that in online racing, where everybody is driving wildly different vehicles?

Leanne Loombe: With the Speed List that you get served up there are car restrictions, so you do have to go in particular cars. We try to get that balance right. We have a wide range of cars, mainly for multiplayer, so we try to balance them in each event’s restrictions.

games.on.net: In a lot of racing games, after a couple of weeks the community has found THE fastest car for each category, and that’s what everybody races. How are you tackling this issue so that your entire collection of cars get used?

Leanne Loombe: That’s part of the reason why we’ve made all the cars available at the start. Then people can go and drive their favourite car.

games.on.net: Cop chases are out of multiplayer. Given that was such an amazing feature of Hot Pursuit, why did you remove them?

Leanne Loombe: Most Wanted has a different focus to Hot Pursuit. Hot Pursuit was very much about racers and cops, whereas Most Wanted is about racing with friends, driving in the open world. Obviously it has chasing with cops in single player, but we do other things in multiplayer instead.

games.on.net: What’s the level of involvement of the cops in single player – will they use all the same tricks that we saw in Hot Pursuit?

Leanne Loombe: Fully integrated – if you start driving dangerously, or too fast, then the cops will be on you. You have to get out of their line of sight, which drops your heat level. As your heat levels increase, more cops will come, and they’ll do roadblocks, and drop spike traps, but no helicopters this time around. You then use the open world to evade the pursuit.

games.on.net: Can you describe the matchmaking, and how that works?

Leanne Loombe: We try and keep the players in the world at all times, so we don’t have any lobbies or anything like that. We do it all from in the world. The matchmaking will happen while you’re in free drive, and then drop you into the speed lists. To find friends you can bring up “Easy Drive” which will show your list of friends, and you can select your friend. You can also create a private or public and match make that way.

Thanks to Leanne for chatting to us!

Let’s be honest here – Need for Speed Most Wanted might as well be called Need for Speed Takedown Paradise. Developer Criterion tuned its racing credentials on the open roads of Paradise City in the Burnout series, and cruising through the streets of Need for Speed Most Wanted feels like a scenic homage to that beloved urban racing arena. It’s a surprising move considering Criterion’s debut Need for Speed title, Hot Pursuit, shifted from the sandbox to a series of lengthy linear tracks, and the pay-off was a tighter focus on more controlled, action-packed racing. Has Most Wanted’s return to open word racing sacrificed the intensity and detail afforded by Hot Pursuit’s linear tracks, or has Criterion managed to bring the best of both worlds?

While I would have loved to test the PC version of the game, PlayStation 3 preview code was the only version available at the event. It was a surprising choice of platform for EA, as most punters will agree that the PS3 tends to run cross platform titles like a 1982 Corolla firing on three cylinders, especially at the preview stage. Unfortunately the demo confirmed these fears, with a framerate that often looked like a PowerPoint slideshow. A bad case of the jaggies didn’t help the game’s cause either, with anti-aliasing undoubtedly sacrificed at the altar of draw distance.

Despite these flaws, the actual geometry of the world still shone through the blocky pixels, with the city environs packed with buildings, bridges and other urban scenery providing a tasty backdrop for the high-speed shenanigans. Car models were also exquisitely rendered, easily the equal of those in Hot Pursuit and lathered in the detail and attention these modern masterpieces of industrial design deserve.

The demo focused exclusively on multiplayer, where eight game writers were pitted against each other in the new Speed List mode. This strings together five different event types, allowing for lengthy online race sessions uninterrupted by menus and loading screens. It’s a style that Codemasters could do well to borrow, given its love of making racers wait. The event kicked off with a speed trap event, where each car was clocked going through a speed camera; the fastest speed recorded wins the event. It was here that I first noticed the odd way in which Most Wanted handles starting grids – there aren’t any.

Each racer is given a point on the GPS that they must head to; only once every racer is in the general vicinity can the event start. A short timer counts down and then the event simply kicks off, regardless of which direction you’re facing or how close you are to the starting line; cars aren’t placed into a neat, orderly starting pattern, which played havoc with my OCD.

It also meant those unfamiliar with the system were just leaving the starting line while the pros reached the half-way point, but it shouldn’t be much of an issue once gamers get used to it.

The next race was a standard point to point event, and it was here that I could really sink my teeth into the new handling model. Given that each car’s stats are tuned to reflect their real life counterpart, the sluggish, heavily-laden feel of my Aston Martin Vantage V12 was more indicative of the real car’s handling than the game’s overall feel. As expected, tapping the brakes before accelerating kicks the car into a drift, while the e-brake (handbrake for us Aussies) should only be used for extreme hairpin makeovers.

Unfortunately I couldn’t test any other cars, so it’s impossible to judge whether all of the cars feel so boat-like. According to the producer they’re not, and based on my prior Burnout/Hot Pursuit experience I’ve got every reason to believe her. My first impressions are that Most Wanted’s handling is definitely more realistic than the zippy radio controlled cars of Burnout, but by no means does it come close to simulator territory.

A novelty jumping event was up next, where racers had to launch over a large ramp to see who could hurl several tons of steel and rubber through the air the furthest. Like all events, it’s possible to sabotage other racers’ attempts by giving them a close encounter of the head-on kind. Smashing a rival out of action is called a Takedown — a name I’m sure I’ve heard somewhere before. Performing a Takedown seemed a little too easy, with even relatively low-speed shunts forcing a car to respawn, losing precious seconds.

By the end of the Speed List I’d also sampled a spot of team racing (where I managed to come second despite being grouped with the Brenna “L-Plates” Hillier) and a drift challenge, which turned into more of a Takedown trial given the intense rivalry between competing gaming critics. Despite the technical shortcomings of the PS3 version, the racing remained fast and furious throughout, and the diverse neighbourhoods of the city always provided a unique backdrop for each challenge.

Any concerns that drivers will simply be hooning through an endless Sydney CBD lookalike can be dispelled, as my brief fling saw me sliding through sewers, careening over construction sights and hurtling down hillside vistas. Whether or not Most Wanted can match the intensity of the police chases in Hot Pursuit remains to be seen, but with a release date just five weeks away we don’t have to wait long to find out.

As well-informed games.on.net readers, I don’t need to espouse the virtues of Solid State Drives to you. By now you’re all well aware that these little blocks of memory can crank your system’s responsiveness to speeds that make a hyperactive ninja look slothful, faster even than a costly CPU upgrade. Even better, they can slash game load times, ensuring you’re always first to get the gunner’s seat in BF3’s choppers when a new map loads. It’s no wonder SSDs were one of the most exciting new PC technologies of the last few years… until the great SandForce Singularity of 2011.

SandForce is a relative newcomer to the tech world, and in 2009 it decided to focus on designing SSD controllers, the chips inside an SSD that figure out how to read and write to the memory modules. Saying that SandForce did an ok job of making these controllers is akin to saying Usain Bolt is a pretty good jogger; SandForce’s controllers wiped the floor with the competition. By mid-2011 nearly every SSD on the market used SandForce’s chips and, to nobody’s surprise, they all performed very similarly. To this day most SSDs still use a SandForce controller (as seen in the SF-2281 that featured in Tim’s recent review of the SanDisk Extreme), which is why Corsair’s shiny new, SandForce-free Neutron and Neutron GTX drives are worth a closer look.

SandForce’s venerable SF-2281 controller is gone, replaced by a controller manufactured by Link_A_Media (LAMD). Yeah, I’ve never heard of this company either. It turns out that LAMD have been making controllers for almost a decade, but it has mainly focused on the Enterprise market. No, they’re not making hard drives for the United Federation of Planets’ Starships; they’ve been building storage for big business, for use in datacentres and the like. With the release of Corsair’s new SSDs, LAMD is chasing the mainstream market, and I’m guessing they’re pretty good at it. Why else would memory manufacturer Hynix decide to acquire the company back in June? I don’t need to guess any more though, as this week Corsair sent me two of their Neutron drives packing LAMD heat for testing; the mainstream Neutron and its souped-up big brother, the Neutron GTX.

Despite the price differences, both drives use the exact same controller in the form of the LAMD LM87800, and they’re the first SSDs on the market to use it. This is an eight-channel SATA 3 6Gbps controller, a configuration shared by most SSD controllers. Both drives also share the same 256MB cache of Samsung DDR2-800 memory, as well as a total amount of 256GB of memory modules. As expected, not all of this memory is usable, with some set aside for drive wear and garbage collection. Corsair sells both drives as 240GB models, but Windows only reports 223GB of usable space once the format fairies have done their job; Corsair aren’t the only manufacturers guilty of this slight capacity exaggeration. Unlike SandForce’s controllers, the LM87800 doesn’t use sneaky compression techniques to boost performance, so it should perform the same regardless of whether you’re working with .zip or .mkv files. ASIO agents might find the controller’s lack of AES encryption a worry, but those of us who don’t have state secrets to protect have nothing to fear.

Where the two drives differ is in their choice of solid state memory. The mainstream Neutron uses sixteen 16GB modules of Micron synchronous ONFi NAND memory (part number 29F128G08CFAAB), which is the same memory type used in Corsair’s previous Force GT drives. This 25 nanometre memory is rated to handle 3000 program/erase cycles before kicking the silicon bucket, and somebody much smarter than I has calculated that this equals around ten years of average use. The more expensive Neutron GTX uses pricier memory in the form eight 32GB modules of Toshiba toggle-mode MLC NAND (part number TH58TEG8D2HBA8C). Once again this isn’t the first time Corsair has used this memory type, with the same modules featuring in Corsair’s Performance Pro drive, and it has the same 3000 p/e cycles as the Neutron’s Micron memory.

With the cost of flash memory making up the lion’s share of an SSD’s pricing, it’s obvious that both Neutrons will have different prices thanks to their differing memory types. At the time of writing, the average street price for the Neutron 240GB is around $245 ($1.09 per GB), with the Neutron GTX 240GB slightly more expensive at $290 ($1.30 per GB). Let’s take a look at how they perform to see if this price difference is justified.

All testing was conducted on an ASUS Maximus V motherboard, using Intel’s Z77’s SATA 3 controller to spoon feed data to the drives. An Intel Core i5-3470 CPU and 8GB of Kingston HyperX DDR3 1800MHz rounded out the hardware, running Windows 7 Ultimate 64-bit edition with Service Pack 1. The OCZ Vertex 4 256GB and ADATA SX900 240GB drives were used for comparative purposes, as both are rated as some of the fastest SSDs currently available.

Our first benchmark was Anvil’s Storage Utilities, rated amongst SSD enthusiasts as one of the best synthetic tests available. Unlike many other SSD benchmarks, it uses data that is hard to compress, which puts the controller’s compression technique through a more realistic workload. The sequential 4MB test is used to indicate performance when handling large files, and both drives performed very well at the read test. The GTX version had a huge 33% performance lead when it came to writing thanks to its toggle memory; while it didn’t take the lead in both tests, overall it was the best performer of the lot when both read and write were taken into account.

The next Anvils test shows 4K performance, which is an indicator of how the drives handle small files, such as those commonly loaded during normal desktop operation. Both Corsair drivers were behind the pack, showing that Corsair has focused on large file performance at the expense of smaller file performance.

Our CrystalDiskMark 4K QD32 is a very similar test, and once again the OCZ Vertex4 took the lead. Surprisingly the plain-Jane Neutron was faster than its more expensive sibling in this test.

We then moved on to PCMark Vantage’s HDD game test, which should indicate SSD performance while loading games. Once again the Neutron took the lead over the GTX, showing the price the GTX has had to pay to achieve such high sequential write performance.

Our final test was the popular AS SSD synthetic benchmark, where we focused on sequential performance. The Neutron GTX nipped at the heels of the OCZ Vertex 4, while the Neutron once again showed significantly slower write performance. Overall performance for both drives is very respectable, with the emphasis on handling larger file sizes, where the GTX is currently the fastest SSD I’ve tested when both reading and writing are taken into account. The biggest surprise is the Neutron’s ability to keep pace with the GTX version in most areas, even surpassing it in some tests.

It’s obvious that the new LAMD controller has the potential to show SandForce a thing or two, and we can expect even better performance as the firmware matures. While the OCZ Vertex 4 offers better value for money at the moment, prices on Corsair’s brand new products will undoubtedly decrease over the next few months. Once prices have stabilised and the drive’s potential has been maximised courtesy of firmware updates, Corsair will have a couple of very potent performers on its hands, and the SandForce stalemate should finally come to an end.

Good:

• Blistering sequential performance
• Decent value for money
• Doesn’t use compression to boost performance

Bad:

• Smaller file performance could be better

It’s not often that I get to say this, but it’s never been a better time to be a PC gamer. Today’s console dinosaurs have both feet in the silicon grave, leaving the PC to reclaim its place as the enthusiast gamer’s platform of choice. These aren’t merely the claims of a PC gamer desperately trying to justify a maxed-out credit card to his disapproving spouse; even EA recently acknowledged that the PC was the second biggest platform over the last financial quarter, not to mention the fastest growing.

It’s not just about the popularity of PC games though; we’re also spoilt for choice when it comes to hardware. The latest round in the ongoing NVIDIA versus AMD slug-fest has been closer than ever, with both companies trading blows over each successive GPU release. AMD delivered an opening haymaker with the powerful Radeon HD 7970 and 7950 back in December 2011, delivering oodles of performance at the cost of watt-guzzling energy needs. It took a few months for NVIDIA to get back on its feet with a technically perfect uppercut in the Kepler-based GTX 680 and 670 products, exceeding AMD’s performance while also delivering products that didn’t need a fusion reactor for power. AMD’s response? Slashing and burning prices, making the choice of which high-end card to buy harder than ever.

This brings us to today’s magical moment, the release of a new product that could not have existed without such close competition between the two GPU giants. The GeForce GTX 660 Ti is NVIDIA’s offering to those who demand the impossible – high performance at mainstream prices. Once again NVIDIA has wheeled out the impressive Kepler design to do the polygon pushing, but unlike the previous GeForce GTX 560 Ti it hasn’t taken a silicon sledgehammer to the card’s ankles. As you’ll see, the GTX 660 Ti has barely been touched by the engineer’s laser scalpel.

Kepler – the nitty gritty

At the heart of the GTX 660 Ti is the Kepler architecture. First announced in 2010, this design was largely a result of the PC industry’s new love affair with mobility. While core gamers crave the desktop behemoths that make up a tiny percentage of NVIDIA’s business, laptops are far bigger slice of the pie.

NVIDIA’s prior GPU design, Fermi, was a gas-guzzling brick that doubled as a cosy foot heater during the winter months, making it a laptop-melting monster. The new GPU design needed to be as comfortable ripping up Battlefield 3 in a desktop as it was in a laptop. Accordingly, NVIDIA’s biggest goal for Kepler was to improve performance per watt — and the company used a variety of means to arrive at its new eco-friendly destination.

First and foremost was the move to a smaller manufacturing process; smaller transistors need less power to operate and also pump out less heat in the process. Fermi’s transistors were built on a 40 nanometre process, while Kepler adopted a smaller 28 nanometre process. Take a peek inside the Radeon 7970’s GPU and you’ll find the same 28nm process at work, yet AMD’s latest cards pump out exorbitant levels of heat by comparison. Obviously something else helped drop NVIDIA’s electricity bills, and that’s where Kepler’s CUDA cores come into play.

Most of the transistors inside Kepler are grouped into units called CUDA cores, and they’re the worker bees in this silicon hive. To lower each CUDA core’s energy consumption, Kepler runs them at the same speed as the graphics clock, whereas Fermi ran them at twice the speed of the graphics clock — but there’s an obvious problem with this approach. Half the speed usually equals half the performance, so NVIDIA came up with a brute-force solution. Where the GTX 580 Fermi had just 512 CUDA cores, the Kepler found in the GTX 680 has a whopping 1536 CUDA cores. These are in turn grouped into eight Streaming Multiprocessor (SMX) units, each comprised of 192 CUDA cores.

All of these SMX units need to be fed data from the graphics card’s on-board memory, and NVIDIA also whipped Kepler’s memory bandwidth into shape. It doubled the total theoretical memory bandwidth of Fermi by using a 256-bit wide bus while also running the memory at a whopping 6GHz memory clock. This gave the GTX 680 a rather impressive 192.3GB/sec of memory bandwidth, which it shared with the GTX 670.

These are just three of the major improvements to Kepler over Fermi, but there are many other smaller improvements too numerous to mention here. The end result was a product that could out-perform AMD’s best, all while using considerably less energy to get the job done. As desktop gamers with a permanent lifeline to the electricity grid, power improvements might not sound too exciting, unless you’ve had to live with a high-end PC that sounds like a hive of wasps every time your GTX 580 kicks into sixth gear. The Kepler purring at the heart of the GTX 680 delivers blistering performance, with whisper-quiet cooler noise as an added bonus.

Pruning time

When it came time to release the cheaper GTX 670, the solution was simple; cut a single SMX out of the 680’s loop. This also gave NVIDIA a way to use chips that had minor flaws in one SMX unit, something that happens surprisingly frequently when you’re building a postage stamp out of more than three billion transistors. The GPU frequency was also dropped slightly, yet the memory bandwidth remained identical. Now, most techies expected NVIDIA to do exactly the same thing with the GTX 660 Ti, disabling one more SMX unit to deliver a product one more rung down the performance ladder from the GTX 670. If the race between AMD and NVIDIA wasn’t so close, it’s quite likely that the GTX 660 Ti would only have had six SMX units, comprised of a total of 1152 CUDA cores.

The good news is that this isn’t the case.

NVIDIA hasn’t touched the number of CUDA cores within the GTX 660 Ti when compared to the GTX 670 – they’re both built using 7 SMX units, for a total of 1344 CUDA Cores. Even better news is that the GTX 660 Ti GPU runs at the exact same speed as the GTX 670, with a base speed of 915MHz. Thanks to NVIDIA’s new GPU Boost technology, when the GPU detects it has thermal headroom to spare the GTX 660 Ti even boosts to the exact same speed as the GTX 670, up to 980MHz. By now you’re probably wondering just what the hell NVIDIA has done to justify selling the GTX 660 Ti for $100 less than the GTX 670?

Instead of touching the number of CUDA cores, NVIDIA has lowered memory bandwidth. While the GTX 680 and 670 use 256-bit memory buses, the GTX 660 Ti drops this to a 192-bit memory bus. However, NVIDIA has kept the same 6GHz memory frequency, theoretically dropping memory bandwidth by just 25%. This should only be a concern for those running stupidly high resolutions or anti-aliasing.

Unfortunately this has been achieved by dropping one of the four ROP partitions that usually make up the 256-bit memory bus. Without getting too technical, the original four-way ROP design plays nicely with the 2GB of on-board memory usually found on GTX 660 Ti cards, but a 192-bit bus is usually associated with either 1.5GB or 3GB of memory. The former is too little for the 660 Ti to be competitive, while the latter is too expensive at this price point (though 3GB GTX 660 Ti products will be possible). Instead NVIDIA is using proprietary technology to force the ill-fitting memory to be friends with the three ROP partitions, technology which it is keeping secret for the time being.

We’re left with a card that is almost identical to the GTX 670, with slightly less memory bandwidth and a recommended retail price $100 cheaper. As you’ll see though, RRPs don’t mean a whole lot in the real world, and the price pressure on every other product has left the GTX 660 Ti in a rather sticky situation. Let’s see why.

To the testbench

We’d like to thank Gigabyte for supplying our review sample in the form of the new Gigabyte GTX 660Ti Windforce 2X. As seen with the release of the GTX 670, most GTX 660 Ti cards will ship pre-overclocked with a faster base and boost frequency, and the Gigabyte GTX 660 Ti is no exception. Base frequency has been given a healthy shot in the arm, increasing to 1032MHz, while Boost speed has also increased, up to 1111MHz. As the name suggests Gigabyte has employed its unique Windforce 2X cooler on this card, while output duties are handled by dual DVI ports and a single DisplayPort and HDMI port. Gigabyte has set a recommended retail price of $399 for this card, but retailers are currently selling it for anywhere between $398 and $450 at the time of writing.

When compared to AMD’s offerings, today’s price puts it somewhere between the price of an AMD Radeon HD 7950 and 7970, the former of which has just received a healthy 15% performance increase courtesy of a new AMD BIOS. It’s also dangerously close in price to the cheapest GTX 670s on the market.

To ensure the testbench wasn’t the limiting factor in the benchmarks, I tested the cards on a machine out of the price range of mere mortals. At its heart is Intel’s rather zippy and ridiculously expensive i7 3960X CPU, mounted in an Intel DX79SI motherboard with 8GB of DDR3-1800. Continuing the Intel theme was the use of a 520 series SSD, while all audio was disabled for testing. Given that this card is aiming at users with 1920 x 1080 displays, all tests were run at this resolution. Ultra detail was selected for every benchmark, and I used the in-game benchmarks for both DiRT 3 and Shogun 2. Battlefield 3 doesn’t include an automatic benchmark, so instead I recorded the minimum and average FPS of the opening 60 seconds of the Operation: Swordbreaker mission with FRAPS.

So, how did the card perform? As expected, 3DMark 11 performance was extremely respectable, almost matching the more expensive Radeon HD 7970. DiRT 3 sadly wasn’t nearly as impressive, with the 660 Ti left eating the dust of the 7970 but proving an even match for the more affordable Radeon HD 7950. Battlefield 3 performance also saw the 660 Ti keeping pace with the 7950, albeit with a better minimum frame rate. The final test of Shogun 2 saw the GTX 660 Ti record a healthy lead over its more expensive brethren, which caused more than a few raised eyebrows. After pinging my local NVIDIA rep, it turns out that this is caused by a performance fix incorporated into the review drivers for the GTX 660 Ti, which should soon make their way into NVIDIA’s WHQL drivers.

The final test used a sound meter to test fan noise while each card was under extreme load, and the Gigabyte card came out in the middle of the pack. Amazingly, Gigabyte’s cooler managed to stay at the exact same volume during my overclocking tests, where I reached a maximum stable GPU core speed of 1250MHz and memory frequency of 7GHz. At these speeds it’s safe to expect another 10% boost in performance, if not a little more.

To buy, or not to buy

There’s no denying that ordinarily the Gigabyte GTX 660 Ti would be a very capable product for the price. But these are no ordinary times, with the fierce competition between AMD and NVIDIA causing video card prices to plummet. At the time of writing, the cheapest GTX 670 in Australia is selling for just $399, while the most affordable Radeon HD 7970 is only $50 more, both prices that the Gigabyte GTX 660 Ti is currently selling for. At these prices both offer noticeably better performance than the GTX 660 Ti, so they obviously get our nod of approval.

Thankfully the shine will start to wear off the GTX 660 Ti’s launch in a month or two, and street prices should drop to around the $350 price point. But even then it’s going to be hard to swallow a 20% performance drop for a mere $50 saving, so Gigabyte is going to have to figure out how to drop the prices even more. As we’ve seen in the past with the GTX 670 versus the GTX 680, NVIDIA (and Gigabyte’s) greatest competition isn’t even AMD – it’s itself.

Good:

• Excellent performance while remaining whisper quiet
• Decent overclocker
• Solid value for money

Bad:

• Outshone by the GTX 670 and Radeon 7970
• Retailers aren’t sticking to the RRP