I noticed someone else realized this will be A HEAVY VOLUME CHIP. They're right. Even at a sightly lower price than X2 3800+ it will be a real seller.The word a few months ago was that it would end up around $125 which seems reasonable since the most expensive Sempron is around $100 and the 3800+ is around $150
Well, according to the linked article, the 3600 just used the X2 512KB cache core.. which makes me believe that it isn't intended to be a high volume chip, at least for now. Once there is a seperate die core (maybe there will be, I don't know), then I can believe this is a high volume chip.
If the X2 3600+ was priced at $125, it would most certainly become a high volume chip. Which would be costly for AMD, especially without a new die. As a low volume product, it wouldn't be cost effective to invest in new masks for an entirely new chip.
The Semprons have a die size of about 82mm^2. This can be mass produced cheaply. A die size of 183mm^2 (the Manchester die size) is not so much so. Even a reduction to a die size of 164mm^2 would probably be worthwhile if it was intended to be a mass volume chip. Then again, because of AMD's DDR pads, it would probably require a complete redesign of certain parts, so maybe that is why AMD doesn't want to make a new die for it (and if that's the case, then it is probably still going to be a high volume chip and you can ignore everything I just mentioned previously).
I don't expect AMD to really drive the X2 (Sempron version) down to the budget market until they have 65nm up and running. Until then, it simply isn't cost effective for AMD to do so. 90nm dual core is simply too expensive. The only reason why Intel can probably pull off the sub-$100 805 is because of their two-die approach, and the fact that otherwise their 90nm fabs would be empty before they transition over to chipsets or a smaller process.