Even so..."consumption" is likely in the 10's of milliwatts range (assuming class 2) no matter how you want to define it, with rolling eyes or not, and not in the whole-watt range as you have surmised. The chip doesn't take that much juice to run. By way of example, a typical usb bluetooth dongle has a maximum *total* power draw of around 400mW or less, and that's with all the logic to run the USB bus on top of the bluetooth logic and radio. In the OMAP3 platform, the bluetooth logic connects via a simple (and by comparison very efficient) serial bus to the chipset. Another example (and perhaps more telling) is that many bluetooth headsets consume on the order of 50mW when actively communicationg (http://www.techonline.com/product/un...2000480?pgno=2) So, no matter how you slice it, the original argument stands insofar that the 3G radio is a FAR larger power hog than is the bluetooth stack.