You may be right in that Tigerton isn't a shrink... It seems that many can't make up their mind as it's projected release is very close to Intel's 45nm date. Intel hasn't officially said either way, only that they will progress to 45nm as soon as possible. Beyond that, we may be looking at it the same way, but focusing on different aspects. Tigerton should most likely be a unified quad-core if it's to scale beyond the current 2-CPU recommended (4 supported, but not with current chipsets) model of the Clovertown. At least that's what Intel said at IDC -- not specifically mentioning Tigerton, but saying that to scale to a significant MP configuration or beyond 4 CPUs, simply placing two dual-core chips on a single die isn't going to cut it.
You have any info on Dunnington at all? Who are you? I'd love to believe you, but... you'd better link.
Right now anything on Dunnington is speculation... But its placement on the roadmap coincides with Intel's projection for 8-core CPUs. They're claiming 8-cores by end of '08 (Dunnington sits on their roadmap as Q4 2008) and 16-cores in 2010. There have been murmurs of 12-core offerings as well, hinting that not all core jumps will be 2-fold. Intel has also pulled Whitefield and Wolfdale from the roadmap and moved other processors up. They're making a bigger push for 45nm and I'd expect 45nm chips on the market in 8 to 10 months rather than sometime in 2008 like Intel was originally saying.
...Anyway, if you can dig up video of the IDC presentation and the Q&A session after, a lot of this was addressed. I don't have anything I can post or link without scanning back through some DVCPROHD tapes a client provided. ...I'm guessing most hardware sites either didn't attend or didn't pay attention because info on the net seems to be all over the place and doesn't even match Intel's own press releases and transcripts from IDC. Then again, just last tuesday, Intel said they are shipping the 120W 2.66GHz Clovertown in limited quantity, yet most hardware sites are still incorrectly reporting that the chip is not yet released. ...Perhaps it's because most of these sites like to pride themselves on "inside sources" and industry connections rather than reporting the official news -- I guess they leave that to the chipmakers themselves and the WSJ.
You can believe any of this or none of this, I really don't care. It's all in constant flux anyway and by the time Dunnington sees the light of day (assuming it ever does), half of the chips on the roadmap that were to come before it will have been cancelled or rescheduled and new ones will pop up. In the tech world, January '09 is still a long, long way off and there's literally dozens of products and revisions scheduled between now and then and a good portion of the info that Intel or AMD release is bogus anyway, just too keep investors optomistic and their competition guessing.
I'm not claiming to have inside knowledge or to be an "expert". I've been in the IT industry for nearly 20 years and have done everything from hardware design to game development. The one constant I've learned is that most hardware sites are typically wrong more than they are right, even though they all tend to spew the same crap. And most people claiming to be an "expert" or a "professional" are usually neither. If you have bothered to read this the whole way through, I commend you, and you won't hurt my feelings if you feel compelled to pour a generous helping of salt on top. I'm just going off of what info I've been able to assimilate and I don't think I'm doing all that bad... Considering Anandtech and Tomshardware are still mostly reporting stuff from before IDC, with a few tidbits of the quad-core launch.