In Theory: How Microsoft could build a new 360 for 2010

A more advanced console with a "built in" Natal also offers intriguing possibilities, not least that the interface between camera and console need not be constrained by USB bandwidth - oodles of it is available for streaming compressed video (eg HD-DVD, or Blu-ray movies) but transferring uncompressed HD data from Natal's precision optics is surely another matter entirely. USB bandwidth at its theoretical maximum would not cope with streaming uncompressed 720p video, not even at Natal's 30FPS scan rate. Natal demos seen in the flesh thus far have shied away from displaying actual camera images on-screen in real time, suggesting that only processed data is sent along the USB cable. If HD video were to be transmitted, it would require compression, the processing power for which would make Natal more expensive. A new interface would lift all of these restrictions. As it is, the bandwidth, data transfer and imaging questions are still something of an unknown - something we hope to address in an interview with Natal personnel at some point in the future.

The more you consider the technical "what ifs", the more an upgraded Natal Xbox could work in line with the original 1UP story, and Microsoft's denials may well simply be a case of the company attempting to safeguard sales of its current hardware. Microsoft doesn't promote new revisions of the Xbox 360 hardware - it was almost impossible for us to get representatives to acknowledge Jasper's existence on the record, let alone get a review sample. As of right now, the message from Xbox mullahs remains constant and clear: no new Xbox 360, Natal will work just fine on the current one.

So having set up the possibilities of the "new" Xbox offering improved technical performance in line with what the original 1UP story suggests, I can still only focus on the doubts expressed in my original blog post. Moving beyond the fragmented marketplace that a "new" console would create, plus the additional headache for developers, the extra RAM/overclocked chipset argument falls foul of two very important factors. Firstly, from an enthusiast's perspective, while the chipset overclock sounds really rather cool, and eminently doable, the fact is that Microsoft has spent years manufacturing out the RROD issues inherent in the 360 that have plagued the reliability and the reputation of the platform.

A 20 per cent overclock uses far more than 20 per cent extra power and typically produces much more heat. This far into the lifecycle of the console, we should be expecting developers to be far more efficient in their coding - the games should be more than 20 per cent "faster" for want of a better term simply owing to them having more experience with the architecture. And in real terms, what does the extra 20 per cent actually buy you in terms of the gaming experience? Take a look at the GTA IV performance video analysis on the Digital Foundry blog. There you're seeing that the Xbox 360 version of the game often runs with a 20 per cent performance edge when compared with PS3, with a 20 per cent resolution boost to boot, more or less. The thing is, you're not getting anywhere near to 20 per cent of a "better" game.

In terms of the memory element, Microsoft does have some form in bowing to the wishes of RAM-hungry developers. It is said that the 360's RAM pool was boosted from 256MB to 512MB on the basis of a presentation from Epic, showing how badly affected Gears of War would have been with less memory to utilise. By today's estimates, an additional 512MB of RAM would cost a mere USD 6 extra per unit. But assuming that Microsoft conservatively shifts another 25 million units across the lifecycle, that translates into an additional USD 150 million over the course. Will higher-quality textures or shorter loading times recoup that investment? Will it make the existing users upgrade their older Xboxes? Unlikely. In terms of the competition, Xbox 360 already runs with a memory advantage over PlayStation 3, factoring out another element that may have persuaded Microsoft to boost the 360's spec.

All of which leads to the inescapable conclusion that Aaron Greenberg is right. It's the Natal hardware itself with all the rich potential it has for creating new gameplay experiences that is the true upgrade here - and the only element that really counts. From speaking with developers, we know that Microsoft's third-party biz team has been touring and showing them the kit in action. So it's not just the word from E3 that is saying that Natal is the real deal, the feedback from the game-makers themselves is that it works.

But where I think 1UP has it right is that Microsoft and Sony will almost certainly be investigating the Wii hardware model for their next consoles in that there's a strong argument for making the new hardware an off-shoot of the existing designs. In Sony's case, by taking the existing PS3 architecture, boosting Cell with more SPU satellite processors and a stronger main core, and replacing RSX with a more powerful, but backwardly compatible GPU, it would have an extremely capable "new" system that developers would already be au fait with from day one. And crucially, it wouldn't cost the earth to launch.

We've already seen demos for ultra-resolution and 240FPS gameplay from Sony created by networking multiple PS3s together. Stereoscopic 3D demos for GT5, Motorstorm and WipEout HD at CES 2009 also suggest that Sony is testing the next generation right now by effectively taping consoles together - in this case by presumably having individual PS3s rendering an image for each eye.

The real challenge will be in how to exact the most gaming impact from the additional tech. TeamXbox reckons that stereoscopic 3D will be the key component in a Natal-equipped true next generation Xbox. While it's fun to speculate on what Microsoft has planned in the next few years, the focus in the here and now is on Natal and more clues as to its potential are beginning to find their way online. Remember the guy who created the Wii sensor bar head-tracking videos on YouTube? If you've not seen them, their relevance is far more apparent now, especially as Johnny Chung Lee is working with Microsoft on Natal and has some interesting insights on his blog...