They hav 3 cameras and will add most probably another 2-3 in the next 12 month all based on the same set of tools:

2 base bodies
EXPEED processor with More or less the same software
51 point AF
LCD screen

By mixing up differently and reusing the technology they can bring up new cams easily and with less R&D cost/unit.

Whatever sensor they introduce they can use it in at least two cameras and two market segments.

They gave up on differentiation but gained flexibility. Current set of cams D300/D700/D3 share so much and are all top in their segments. Just a new sensor would add two other models.

Modularity does not mean that you can exchange YOURSELF a part, it means you can benefit from same technology in various combinations.

I work for a big IT provider and we do exactly the same in our products.

1) 24mp FX
2) 12mp FX
3) 12mp DX

This gives you options for high resolution FX (without high ISO capability), high ISO capability (without high resolution capability), and cropped sensor for telephoto applications (without high ISO capability or high resolution capability).

So you could own one camera and install the right sensor for the right application. This should be cheaper and handier than owning 3 different DSLRs.

When you figure many pros will shoot for all 3 scenarios and want at least 1 backup camera, this really makes sense. It should be cheaper to have 2 modular bodies than to buy several specialized bodies.

Modularity allows one body to overcome the physical size limitations imposed on the photosites. I see this as a D3 style body that takes sensors in 24, 12 and 6 mp sizes. The processor would be speced to the 24 mp chip so it would provide even better perfomance at the small sizes. But the 12 and 6 mp chips would offer radically better ISO performance than the 24, and would provide substantial improvement in the overall function of the camera. So a shooter could take one body to a location, and when the sun went down, switch to 6 mp sensor and be ready to go.

I think that fully satisfies criteria 2, and is probably what Nikon is thinking. After all the CPU is largely irrelevant to this discussion. On the pro cameras, size of the body tends to be the limiting factor for the CPU, and there is no real trade off, a better CPU is simply better, while 24 mp is better for some tasks, and 6 mp is better for others. So the CPU in the D3x is up to the task on smaller sensors. But smaller sensors, because of the physical size of the photosites, provide a function that the larger chip simply can’t.

Right?

The cubic modular ist the better way