45nm is the way to go
As 65nm Phenom mask would be bigger than the one that AMD has with K8 at 65nm some important people have decided to canceled the dual core Phenom until 45nm arrives.
The decision was business and strategically as AMD likes the way that dual core Athlon performs and these dual core 3.2GHz CPUs are selling for under €100 as we speak.
It would be very tough to position dual core Phenom and AMD probably has the same clocking problem with these dual cores.
Dual core Phenom in 45nm should be scheduled for Q2 2009 and from this time point, this looks awful far away.
I think this was pretty much a given. K10 just isn't clocking high enough on a 65nm process to really warrant replacing the K8-based chips (which are cheaper to make, probably a big factor in this decision). Ideally, this means that more time and resources are being put in to the 45nm K10.5 chips.
The other factor here is that I think ultimately AMD wants to push tri-cores into dual-core pricing territory. This would allow them to compete in the higher end market (with 4+ cores) and the mid-range market (3 cores) with the same die, which would lower manufacturing costs quite a bit. In addition, doing this would essentially push dual cores back into the super-low budget category at which point it matters little whether or not it's a K8 or K10 core behind it.