Here comes the Microsoft Xbox 3 / Xbox 720

Game consoles are just specialist computers and so, like all computers, they obey Moore’s Law. This is something that it is very important to keep in mind when looking at the console generations. Moore’s Law states that the number of transistors that can be placed on a chip doubles every two years. So if a new console is launched 6 years after its predecessor it will be 8 times more powerful. This also means that a given game console should approximately halve its manufacturing cost every two years. It is this these realities that drive our gaming hardware forward.

Now let’s look at our current generation of HD consoles, the Xbox 360 (first available Nov 2005) and the PS3 (first available November 2006). These are both architecturally very different from their predecessors, the Xbox and the PS2. This meant that the whole industry had a massive learning curve to get games working well on them. We have now climbed that learning curve and are driving these machines to pretty near the max, so in pure processing terms games are not going to get much better on them.

The next generation of home consoles from Microsoft (definitely) and Sony (probably) will be just massively upgraded developments of the current generation. They will simply use the latest evolutions of the same CPU and GPU families. Their introduction will not be industry disruptive because they will be 100% backwards compatible. Not just with games, but also with online services like Xbox Live.

So when will we see these machines? It is simply a matter of choosing when to freeze the specification of the processors. The longer they leave it, the more powerful the machines will be. Moore’s Law. But there are huge advantages of being first to market, as Microsoft proved with the Xbox 360. So it is a matter of balancing commercial reality against computer power. In the past this balance led to new consoles being introduced after approximately 5 years. Which means that a new Xbox could be with us next year (2010).

But now we have Natal to muddy the waters. Natal is a step change in what video gaming can do and puts Microsoft a long way ahead of its competitors (presuming Natal works as advertised). So effectively adding a Natal unit to an existing Xbox 360 gives the user a jump in capabilities comparable with buying a new generation machine.

This brings us to two scenarios. The first is that the Xbox3 / Xbox 720 will be with us next year fitted with Natal as standard and that Microsoft just showed us the Xbox 360 version at E3 to demonstrate the technology. The second scenario is that Natal gives Microsoft breathing space to delay the introduction of a new machine by, say, a year. Thus making the new console significantly more powerful. Certainly Microsoft will have had many internal strategy meetings to work out which of these scenarios works best for them.

Currently the rumour machine is going for the first option and Microsoft aren’t denying it. Which would put the new machine in your local shop in 18 months time. Start saving.


  1. No brand new consoles coming any time soon. Just because they can make faster machines, doesn’t mean that they have to make them if it’s not beneficial to them.

    Sony won’t make a new machine because they need to make money on the one they have already first. Nintendo may make a machine but it will be purely catch up and they don’t care about this fight.

    I agree that Microsoft will release a Nadal version of the 360, slimmed down with the Nadal having it’s own hardware to deal with recognisition, less noise and less red ring death. This goes in line with all machines eventually being slimmed down. But I think there will be no difference between that and having a current 360 and plugging a Nadal in. The little bit more memory and processing power may just make ‘HD’ versions of game viable …but more just to get games actually running 1080p rather than adding more wiz-bang.

    But a whole new machine? They’d be jumping the gun and ending up with a Dreamcast. Then Sony would just have to wait a couple of years and release hardware that is a lot better to directly compete with it (Moore’s Law).

    It’s going to end up like a wild west shoot out with the companies staring at each other until one dares make a move.

  2. “Game consoles are just specialist computers and so, like all computers, they obey Moore’s Law.”

    Utter tosh. Moore’s “Law” is no such thing. It was temporarily an identification of a trend in hardware evolution, perpetuated by the media and the PC/chip sales departments. The transistor count on CPUs hasn’t doubled every year for a long while. The chip manufacturers shifted to upping the clock speed (GHz) of a processor. That in turn has run into the laws of physics, hence the shift towards multiple processors with a slower clock speed.

    Your basic premise is flawed, and everything else here is pure speculation based on that flawed premise. Including the mistaken belief that we are pushing the current generation anywhere near its limits. The PS3 in particular is massively underutilised, but the 360 is only slightly less so; the problem, as it has been since the generation shift, is that the basic economics of generating enough high definition content for this generations titles mean are unprofitable. The barriers for this generation are not technical; and Sony and Microsoft gain absolutely nothing from releasing a new version of their hardware, to reset the learning curve again.

  3. Chris, here is a nice Moore’s Law graph that seems to demonstrate that it works:

    Plus Intel’s own explanation:

    The platform holders will definitely release new consoles. Because of competetive pressure from the other platform holders and new entrants. Apple for instance.

    I would argue about the PS3 being massively underutilised. Every developer I speak to says that the GPU is a massive roadblock.

    And as for generating enough high definition content, the more power the platform has the more you can use middlware. Early games, on 8 bit machines, were in machine code. But with each increment we have ever seen in processing power we have seen an increment in middleware.

  4. While there’s no doubt that the Xbox 360 will see an incremental upgrade, the 2010 estimate seems to have been pulled out of thin air by 1UP.

    Developers may be running out of headroom with the Xbox 360’s current spec but I would expect there would be a lag (of years, possibly, as we saw with the PS2) before that impacts sales. A new console in 2010 would also limit the window for selling Natal as a premium accessory next Christmas a la Wii Fit.

    Exciting (and expensive) times ahead either way.

  5. 2010 seems likely now that Sony have come out and said that their motion detection stuff is coming out in Spring 2010.

    I don’t think that Natal will be that early, but knowing Sony their motion detection stuff will come out in Japan first giving Microsoft a few months spare to work on a global release.

  6. s3ntient,
    One rumor does not make another rumor wrong. They are all just rumors.
    In my piece I tried to explain the impact of Moore’s law on the console generations.
    As for 3D, I have written about it on here and seen it in action. However Microsoft cannot limit the Xbox 3 to just 3D televisions, otherwise they wouldn’t sell very many. So while it may be 3D capable it will also work fine with good old 2D televisions.

  7. Hi Bruce,

    Sorry for the short comment, however I actually went into more detail as to why I think you’re wrong on your blog on

    With regard to the article I linked, it was only for the MS rebuttal to the rumor you based your article on. Personally, I don’t think they will be releasing a 3D xbox in the next couple of years for the same reasons I stated on your other blog.

  8. I’m guessing it as this; Slim ps3 on TGS, Microsoft announcing its next full fledged console on E3 2010 and launching on christmas 2010 just to tease Sony and try to force a new console from them.

    If I were Sony,I would retire for the whole 8th generation and then reappear at the ninth.

  9. Sony could release a minor bumped up console, similar to what Nintendo did with the Wii. So the PS4 could contain 1024 megs of shared RAM and a new slightly increased GPU. Memory is cheaper now and the IBM roadmap calls for a Cell with 4 PPU and 32 SPU. But that new cell might be cost prohibitive. So stick with the current cell, albeit in 32nm, and enjoy the cost savings. But the better GPU and much more memory would be enough for developers to exploit for another generation (another cheaper gen for Sony). And it would make the PS4 fully backwards compatible as well.

Comments are closed.