Jun 18

The end is nigh for the modern graphics chip. It genuinely pains me to say that. After all, I’m an unapologetic chip aficionado, someone who loves the technology of integrated circuits for the sake of it. But it’s becoming increasingly apparent that GPUs are over-engineered, increasingly irrelevant and almost definitely not long for this world.

The background here involves a confluence of technological trends. The most ominous of these in terms of the GPU’s longevity as a discrete component is the architectural convergence of CPUs and GPUs. However, one of the most debilitating symptoms of the graphics chip’s terminal malaise is complexity – sheer, pointless complexity. Take Nvidia’s latest uber pixel pumper, the GeForce GTX 480. It weighs in at three billion transistors. That’s getting on for triple the size of Intel’s beefiest PC processor, the six-core Core i7-980X.

If the GTX 480 was any use, that monster transistor count would actually add to the allure. But the harsh truth is that it isn’t – for almost anything. And that makes it dumb. You see, despite the hype regarding running non-graphics applications on GPUs, there’s still very little outside of games that makes more than passing use of a desktop or laptop GPU. More to the point, the number of games demanding a really high-end GPU that are actually worth playing isn’t merely a small number. It’s zero.

Put it all together and you have a terminal mismatch between the cost and complexity of GPUs and their real-world utility. In truth, I’ve felt this way for some time. But it’s the apparent emergence of a radical alternative to established 3D rendering technologies that really brings home how bloated and ludicrous graphics chips have become.

This alleged revolution in rendering comes from a small Australian software startup known as Unlimited Detail. It’s not actually brand spanking new, having been in development for a year or three. But thanks to the random nature of web-based content aggregators, Unlimited Detail was lifted from obscurity recently in a flurry of YouTube-powered publicity.

Anyway, as far as I could tell the basics of this new rendering technology involve ditching polygons in favour of atomic points in 3D space. The claimed result is quite literally unlimited geometric detail. Oh, and the whole thing runs in software at smooth framerates on a conventional PC processor. The GPU doesn’t get a look in until it’s time to spit out the final 2D images.

You hardly need me to point out it all seems too good to be true. So, there was nothing for it other than to go straight to the source and speak to the guys at Unlimited Detail. The technical brains are provided by Bruce Dell, a former supermarket manager, while the business nous comes courtesy of Greg Douglas, a games insider formerly of developers Auran.

The idea of using atoms or points is not new, of course. The really clever bit in UD is the 3D search algorithm developed by Dell. The precise details are UD’s big secret. But according to Dell, “The algorithm takes point cloud data and files it in a certain way so that it can be quickly sorted and accessed.”

When the algorithm searches for points, it doesn’t do so indiscriminately. Instead, it only pulls up a single point for each on-screen pixel being rendered. “We only grab the atoms we need for each pixel, we don’t touch the others,” explains Dell. In other words, the workload depends on screen resolution, not the underlying geometric detail of the scene being rendered. Thus, an impression of unlimited geometry is created.

The UD guys claim the algorithm is so efficient it runs in real-time in a single thread on just one core of a conventional PC processor. Apparently, it will even scale down to simple CPUs in mobile devices.

So far, the only hard evidence for these incredible claims takes the form of a few pre-recorded videos of dubious quality. However, having spoken to the UD pair, I’m happy to confirm they’re not only incredibly passionate, but strike me as completely genuine. It’s potentially extremely exciting stuff.

Still, even if UD works exactly as advertised, the established players in graphics are hardly going to embrace a technology that instantly renders several decades and billions of dollars of investment obsolete overnight. You have to assume Nvidia, and to a lesser extent AMD, will resist the idea strongly. But if Unlimited Detail’s technology gains any traction at all, GPUs really will look sillier than ever.

Tags: amd, application, business, circuit, convergence, CPU, desktop, developers, Development, device, ims, laptop, Mobile Devices, processor, rms, Software, space, Technology, transistor, web, XP, youtube
Jun 11

This model was created by Andrew Comb. It took six years to model in Lightwave 3D, and has a total of nine million polygons.

Once upon a time, a handful of pixels madeth the space invader. Graphics were iconic, not representative: a picture on the box or manual showed you what it was meant to look like, and your mind filled in the necessary gaps. Nobody could have predicted that in just 20 years, we’d be immersing ourselves in realistic living cities, flying over gorgeous tropical islands and going head-to-head with astoundingly rendered characters – and not even being particularly impressed. But in years to come, modern games like Grand Theft Auto IV and Crysis will look just as dated as the classics we remember from the days of yore. In fact, they’ll probably look more so: while the simplicity of a retro game’s look has a certain charm to it, old 3D titles tend to flat-out look old. Try almost any hit game of the mid-to-late ’90s for proof of that.

3D is about more than just pretty graphics. Done right, it makes gameworlds come alive. A 2D sprite can only do what it’s been drawn to do, while a 3D character has a complete endoskeleton and can respond naturally (at least in theory) to anything that happens – the classic example being ‘ragdolls’, where a fallen enemy doesn’t simply slump to the ground in a canned animation, but tumbles off the railing and lands with one arm draped over a step.

You can create worlds rather than merely levels, unlocking the player’s ability to truly explore and experience the world as the character would. You can build simulations ready to be poked and prodded, abused and enjoyed.

It’s phenomenally powerful, to the point that many nominally 2D games are now really 3D ones viewed from a locked perspective, so that they can better use the possibilities of animation, physics and art assets. Why render hundreds of frames of animation you may not be happy with when you can make a model and keep tweaking it until it’s perfect? You may lose some of the old-school charm, but you gain far more.

Dawn of a dimension

The first big 3D success was Battlezone, a tank game released in 1980 that used vector graphics to create its work, much like Asteroids. While a simple game by modern standards, it was fiendishly complex for such an early example, offering the ability to go anywhere in an (admittedly featureless) world, hide from attacks and fight enemies.

Battlezone was thought to be so realistic that the US Army used it to train tank gunners.

Not impressive enough? In 1987, the first Freescape game, Driller, hit the shelves. It offered a full 3D world on platforms as basic as the Spectrum, and was an actual game rather than just a tech demo. Next to that, it didn’t matter that it was ugly, the frame-rate was abysmal and the game itself wasn’t actually much fun – it got lots of attention.

The Freescape engine in its various forms was used in several famous releases, including Castle Master and its sequel, and the all-out 3D Construction Kit. Legend has it that someone somewhere once made something other than a surreal, unplayable mess in this, but we never saw it. Freescape also made it onto TV, in the form of the absolutely atrocious Craig Charles vehicle Cyberzone, one of the most toe-curling attempts at creating a games-related TV show ever. Thankfully, all that remains of it is a single YouTube clip – and that’s painful enough. Most of the early 3D games stuck to simpler technologies. These days, we think of 3D as free-roaming, real-time engines, but back in the day, simply getting a game to look 3D was impressive. As early as 1981, games were achieving this feat – 3D Monster Maze was terrifying a generation with its slowly updating screens and roaming T-Rex, and the first Ultima game was offering a very advanced hybrid of block-by-block movement 3D graphics for its dark dungeons alongside a top-
down 2D overworld for exploration.

Interestingly, while most developers kept pushing further towards 3D, Ultima ended up pulling back, switching entirely to a top-down sprite-based system for the series’ glory days. The 3D element later developed into its own spin-off – the Ultima Underworld games – before returning for the series’ sadly disappointing final outing – the buggy, system-murdering Ultima IX: Ascension.

Faking it

The problem has always been the same: the potential of 3D fights with the limitations of current systems, whether it’s simply displaying the graphics in the first place or making them look as good as other art styles. Going back to an early ’90s 3D game now is almost painful. Flat faces, non-moving lips during conversations, stick-figure character models, smeary textures and appalling animation… the list of problems goes on. Some games got past this, sometimes bizarrely – most obviously Core Design’s legendary heroine Lara Croft, who managed to become an international sex symbol despite looking like a pointy-chested Pinocchio. Most survived simply because playing a 3D game felt futuristic, even if the lack of polygons our PCs could push out meant that 2D games were usually much more detailed.

For much of 3D’s history, the trick has been getting the effect of the third dimension without having to do it for real. The early Wing Commander games gave the illusion that you were flying through 3D space, but really they were just scaling sprites up and down. In first-person shooters, it quickly became clear that walls were easy thanks to their incredibly simple geometry, but snarling hellbeasts dripping blood from their fangs were asking a bit too much. So developers compromised. The worlds themselves were made in 3D, initially just as mazes. Then, as texturing became more advanced, more realistic-looking areas were created, like those in Catacomb 3D.

Interestingly, the state of the art varied dramatically across genres. Shooters had to be fast and fluid, so they were kept simple. In the case of early games like Core Design’s Corporation, things were stripped down so much that the engine didn’t even bother with textures. Id’s first breakout hit – Wolfenstein 3D in 1992 – had textures to depict the inside of its supposed Nazi castle stronghold, but all the maps were completely flat and the interaction was limited to just opening doors and shooting enemies. Ultima Underworld, which came out in the same year, had sloped surfaces, advanced lighting effects, dialogue, puzzles, magic systems, physics, 3D objects, a real plot, the ability to look up and down instead of having your view locked straight ahead and much more.

Ultima Underworld could afford to push these limits because as a role-playing game, it was inherently slower than a shooter, and the audience was more willing to accept the necessary limitations, like the small viewing window. It didn’t hurt that while publisher Origin’s official motto was ‘We Create Worlds’, its unofficial credo was ‘Your PC Will Cry’. It never did worry about system requirements…

Two and a half dimensions

For performance and sales reasons, most games were stuck in what was dubbed 2.5D until the launch of Quake in 1996. Quake wasn’t the first genuinely full-3D shooter in the modern style, but it was the one that made 2.5D officially obsolete. These faux 3D or 2.5D games were flat maps, where areas could be raised and lowered (and in later games like Duke Nukem 3D, sloped), but you couldn’t have one room on top of another. Some games pretended otherwise, but it was generally a trick – the player would be silently teleported as they crossed the room’s threshold. Descent provided a full 3D world made up of polygons in 1995, but only a basic one – it was a series of sprawling mineshafts. The Star Wars game Dark Forces offered rooms above rooms, but otherwise stuck to the standard technologies on display in any other shooter of its era, including sprites.

First-person shooter bosses like this were unheard of before Quake was released in 1996. Now, they’re standard issue.

Sprites were a growing problem. By the mid ’90s, level design was getting more and more impressive. By modern standards, the opening cinema level in Duke Nukem Forever is empty and unconvincing, but to an audience used to bland military base corridors and castles, the realism was incredible.

Almost all these games were able to do this because they saved their 3D for the world. Actual characters were pre-drawn 2D images, pasted in to the game. Not only did they increasingly look weak, not part of the scenery and incredibly blocky up-close, but they also didn’t fit. Games could scale sprites to deal with the player getting closer to and further away from them, but when they started offering the ability to look up and down (first made popular with Heretic, a 1994 fantasy game) the effect showed its limits, with sprites shearing and increasingly looking like the cardboard cutouts they were.

A little voodoo magic

Games had to get more advanced, but the performance wasn’t there. Quake brought full-3D worlds and enemies to the field in 1996, but at the cost of visuals. Technically, they were better, and the improvement in animation was stunning – an early sequence where a snarling Fiend leaps out of a door to attack the player directly is one of many gamers’ fondest memories, to say nothing of the first episode’s room-
sized, lava-throwing monster Cthon. But the world was still drab, ugly and simplistic, with enemies that looked like they’d been knocked into shape with a sledgehammer and textures that would have lowered the tone in a morgue. Like all graphics technology, this quickly improved over the next few years, but it was increasingly obvious that simply throwing more CPU power at games wasn’t going to cut it.

Not alone, anyway

Most people didn’t really see the need for dedicated video cards at first because, to put it bluntly, there wasn’t much of one. Games increasingly offered a high-end mode that offered a higher resolution and additional effects without a specific need for extra hardware – bar a hefty processor. This high-end mode slowly became the standard experience, yet it was a long time before video cards became mandatory for playing PC games. Even technical showpieces like the original Unreal would run in so-called ‘software mode’. There wasn’t one big game that marked the transition, more a slow giving into inevitability.

Dedicated 3D cards came into their own around 1997, with several competing, mutually incompatible brands fighting it out for dominance. By far the most successful was 3DFX with its Voodoo cards. Even now, with built-in low-end 3D a standard on motherboards, a separate 3D card is mandatory for most games. On the plus side, the dominance of DirectX means that you don’t need to worry too much about which card to buy.

Individually, the abilities offered by 3D cards don’t sound too exciting. In the late ’90s, the core functions were transformation, clipping and lighting – in short, getting the card to work out where items in the world were and how they should be lit. The next big advance was the addition of shaders. Shaders are additional calculations thrown into the rendering pipeline that work on individual pixels, vertices and pieces of geometry to add effects and change the final image. Examples include working out appropriate shadows, making a flag flutter or adding bump mapping (a texture that gives the illusion of raised and lowered areas on an object without the need to add polygons). More advanced shaders include motion blur and bloom effects, soft shadows, depth of field and volumetric lighting.

The new bottlenecks

With the technology burden eased (or at least partly passed onto the card manufacturers), developers could focus on making the most of what they had. Valve largely pioneered skeletal-based characters rather than keyframe-by-keyframe animated enemies in Half-Life, and Ritual went out of its way to create interactive elements such as hackable computers in the real-
world environments of Half-Life’s closest rival of the time, Sin. Ragdolls spread to every new game until the games that didn’t offer them felt stodgy and ancient in comparison. Most importantly, the new realism of these games finally let developers sink their teeth into genres that they’d never have been able to do as traditional corridor shooters. Grand Theft Auto III gave us a living city. A million games brought the horrors of World War II to life in glorious cinematic style.

Realistic character and facial animation is a surprisingly recent addition to gaming.

Problem solved? No, just replaced. Now the issue became one of creating all this in the first place. A simple maze-based 3D shooter could be churned out in under a year by a competent team, but building cities, battlefields and other real-world elements requires an immense number of assets. Games are much shorter than they used to be, not because they’re more complicated – in most cases, the features are more advanced but the thinking is more conservative – because of the amount of content required, and the cost of making it. There are shortcuts for some things, like the Speedtree libraries for procedurally generating a forest in a hurry (used in, among many others, Grand Theft Auto IV, Elder Scrolls: Oblivion and Fallout 3), but in general, if a developer wants something, that developer has to make it himself.

Crysis 2 is the current state of the art: explore New York in the wake of an alien invasion.

Technology itself has also reached a plateau, not because there’s nothing more to do – nobody thinks that – but because right now, the big money is on Xbox 360 and PlayStation 3. Both are relatively old machines, but that doesn’t matter. PC ports often allow for higher resolutions and sometimes let you switch on better graphics (Metro 2033 looks much better for instance, as will Crysis 2 when it lands) but it’s usually a token gesture. To put the graphical difference into context, many PlayStation 3 games don’t even use anti-aliasing to smooth out the edges of their polygons, and games on both platforms typically render at a lower resolution than the high-def numbers you’d expect from the systems’ marketing.

Tags: assets, Computer, computers, context, conversations, CPU, developers, Environment, generation, Hardware, ims, iss, lighting, maps, marketing, money, performance, physics, processor, requirement, rms, sap, sla, Software, space, system, Technology, XP, youtube
Jun 08

There’s always analternative to buying a new PC and that’s to upgrade the components inside your slow machine.

Everyone knows the feeling. Your PC has become slow and unresponsive, and it’s getting rather noisy too. All around you are adverts for fast new machines – PCs groaning with cutting-edge components and fancy new features; machines untouched and factory fresh. It’s easy to wilt under this kind of pressure and give into the new PC dream – but luckily we’re here to help you to renew your willpower and stand firm against such temptations! We have happy news: a few well-targeted and cost-
effective upgrades can transform your flagging PC into the machine of your dreams.

So why has your machine got so slow? Well, just as time can be particularly punishing to Windows’ boot time, so it can ravage the hardware in your system. Dust can clog fans and obstruct airflow, overheating your components and bringing their efficiency down. Add to this physical consequence of time’s passing the fact that in the years your PC has been sitting in your home, clever-clogs developers have continued to plug away at their work, creating faster components than those in your machine were even when they were factory-fresh. We’re sure you can see how the problem has arisen.

Refresh your system

This means the best way to speed up your PC is to give it a spring clean and identify the components worth upgrading. In general terms, a PC will last you for a couple of years before you either have to do a major upgrade or piece together enough smaller ones to make the system continue to be fast enough for everyday use.

Which upgrades are best for you is ultimately defined by what you use your machine for. If you’re into video editing and production then a healthy amount of memory will make moving clips around that much smoother, while a faster processor will render your effects and final edits quicker. An SSD will boost program loading times as well as your OS’s boot time.

Audiophiles have similar needs, but they should also keep an eye on how much noise the system is making. Designers benefit from an overall system refit that focuses on an SSD and embraces the current low in memory pricing. The programmers among you will be hankering after more memory as well, along with access to newer motherboard technologies such as USB 3.0 and SATA 6Gb/s – purely for research, of course.

Gamers will see the biggest boost from a graphics card upgrade, because while the promise of multithreaded gaming is closer than ever, the graphics card is still the biggest bottleneck in most systems. And two years is a long time in the graphics card business – we’ve seen the release of not only affordable DirectX 10 cards in that time, but a new breed of cutting-edge DirectX 11 hardware as well.

Click the links below to discover which components you need to upgrade as we talk you through how to do it, and reveal how a tissue can help speed up existing hardware.

Upgrade Your RAM

Understanding Graphics Cards

Cool Down, Speed Up

Replacing The CPU

Select The Perfect Motherboard

Tags: business, CPU, developers, Hardware, Health, iss, memory, processor, Research, rms, rpc, system, Windows