45nm Quads?
 

I've been trying to find out an expected release date of these and can only find info that says Q4 of 2007 and others say Feb '08 and being Feb. 28 I'm kinda curious when these are due out.

Anyone have an inside scoop? or should I stop holding my breath and just get a e8400?

What's wrong with the G0 Q6600's? They are great quads, run cool and get good clocks.

new quads have the sse4 instruction set, have more cache, and are supposed to be 50% faster as a result

Only on applications specifically optimized for them, which I believe are pretty much limited to encoding/AV/server types. For most applications, it's single-digit percentage improvements. The power savings over the 65nm quads disappear pretty quickly once you start overclocking them.

I would expect the larger cache to be the bigest benefit depending on the application.

Folders should see a boost, and super-pi benchmark score afficionados should be pleased as well.... :)

I ordered a Q9450 on Feb 2nd from NCIX - the ETA was first Feb 20th, then changed to March 3rd. We'll see next week. I'm a patient man.

50% lol, 10% more like it.

They also have lower multi's, meaning a Q9350 won't be easy to get over 3.5GHz, even with water.:blarg:

Well I'm eyeing up a Q9450 but it does seem like a toss up against the q6600 between the better multiplier and the bigger cache and chip improvements. If the new quads like the FSBs that the e8400's like then I'll be happy though I am being realistic and expecting lower given its a quad. But at this time the plan is to get a Q9450 and push for 4GHz...

You'll never get a quad to run the same FSB's as a dual-core, barring some miraculous new development. Even the QX9650's haven't been doing well past about 460 FSB (460*8.5=3.9GHz). Good OC'ers have been getting a little past 470, but you need a really good board to do it, and you're hitting a serious wall by that point.

I've been seeing some talk regarding the supposed errata on Intel's Quad-cores, and some suggestions that a new "C1" stepping will help matters a little, but that's entirely speculative at this point, and quite dependent on how honest Intel's being with this 'errata'.