Here's what I think happened. When AMD was deciding on what chips to make on what process, they weren't sure how the 90nm process would go. They knew that making chips in the 1.8-2.2GHz range should be easy, but had no idea that in fact most of the chips would be capable of 2.6GHz or even higher. As such, they made a decision that their high revanue chips should be made on a process they KNEW was capable of producing 2.4GHz chips, i.e. the 130nm process (BTW, I think all 130nm 939 chips are technically newcastle cores, not clawhammer - can anybody confirm this with a CPU-Z shot?), since they would hate to piss off people spending $700+ on a CPU if it can't reliably run at 2.4GHz. Anyways, it turns out that they could have just transitioned all 939 chips to 90nm, and probably dropped prices and sold a crapload more higher end chips (a lot more people would be considering a $500 4000+, for instance, bringing revanues up, not down). Oh, well, you live, you learn.
Anyways, I think the transition point will be with Rev E0, a.k.a. venice and san diego, due out over the next few months. There is no point to make a winchester core 4000+ when a san diego (or venice, 2.6GHz 512K L2) is gonna be out like a month or 2 later, with SSE3, better mem controller capable of using 4 DDR400 dimms, and Strained Silicon.
On a side note, why do you want to grab a 4000+ if you're overclocking? There really is no incentive to do so, as lower chips can OC just as high. Why not save yourself a few hundred bucks to go towards a sweet video card, better memory, better mobo, etc.
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)
Discussion Intel current and future Lakes & Rapids thread
