To answer that question, yes, Ivy Bridge is limited to quad-core at present. That is more due to the limitation of the LGA 1155 socket that all current Ivy Bridge CPUs use. Hence, it is the limitation (currently) of the LGA 1155 platform, not Ivy Bridge. A hexa- or octo-core Ivy Bridge requires a much more expensive and higher-end socket such as LGA 2011. Unfortunately at present, all of the current LGA 2011 CPUs (including the Xeon E5 series) still use the older 32nm CPU manufacturing process (Ivy Bridge uses the 22nm process).
With the Sandy Bridge and Ivy Bridge generations, Intel has been bringing out these new technologies first on the mainstream platforms rather than the high-end platforms (unlike what Intel did with Nehalem, which was introduced on the high-end platform that very few people could afford at the time of introduction). That way, the middle-level PC users get the newer tech first before the entry-level or high-end users do.
TDP. Microsoft wants to limit the TDP of their chips (see http://nl.wikipedia.org/wiki/Thermal_design_power) and thus has to either limit the clock speed or the number of cores and L3 cache.
You may wonder what all this means and I can understand that. Look at it like a car, where the manufacturer limits fuel consumption to X gallons per mile. He either has to limit the top speed (clock speed) or limit the number of cylinders in the engine block (number of cores). That is what Intel currently does.
For AE and 3D applications, number of cores is critical, for PR clock speed is critical. Since you mention octo-cores, there is only the i5 Xeon with 8 cores. Bulldozer or whatever other AMD 8-cores there are on the market now or in the near future are absolute crap in comparison to Intel.
FYI, I'm waiting for the Ivy E version, 22nm models that lift the 6-core and 15 GB L3 limit of the current SB-E and will give us 8-cores and 20 GB L3 cache. No idea when it will happen however.
Thanks guys! I saw a 2012 intel roadmap and started to get that sandy bridge/e will be ruiling the high end roost for a minute.
Harm, you hinted at something that's been nagging me for a while. I keep hearing a debate go back and forth about ghz vs cores, and some people swear by overclocked i7's while other people say it's more important to have more cores. My primary work is in AE and C4D, and I've been agonizing over lower ghz dual xeons (2.5ghz) vs high end i7's. It sounds like for 3d, AE, and motion graphics work (a lot of particle systems, etc), that a dual xeon 6 core, even at lower clock speeds, is going to be preferably to a high end i7?
I'd of course love to get the fastest best most core xeon, but about 8-900 per cpu is max of my budget if I go dual cpu.
Well, there is no doubt that dual i5 Xeons are faster for AE and 3D work with all those cores, but they come at a price. For PR I still think that an overclocked i7 is faster.
So you need to do the math,
i7-3930K and a P79 mobo for around $ 900 or less, or
dual i5-2690 and a corresponding mobo for around $ 5000,
disregarding the cost of memory and the rest. IMHO there is no question, the i7 delivers a much better bang-for-the-buck and even then a new system will easily cost more than $ 5K, once assembled.