Should be clear why that is, AMD doesn't have the luxury to sell locked chips with 2GHz base clocks as all their chips are unlocked. So while Intel could throw away their less than ideal chips even at uncompetitive prices, as locked variants, in retail or OEM channels. AMD cannot do that, now whether they tried or not is up for debate.Given proper cooling, I don't think there isn't an intel CPU that won't do the boost frequency on all cores if you set it to do that in the bios. That is technically overclocking, but it does work. AMD's boost is like this magical thing that sometimes boosts the cpu higher than we can overclock it ourselves. Intel's boost is deterministic. AMD's is something else.
EDIT: I think it was debauer who said he is unsure if AMD even knows why their boost is behaving the way it is.
Continuing from the other thread, there is indeed a problem with AMD's advertised or rated boost clocks because it shouldn't be this random. It would greatly helpful if they put a reference (spec) PC with workloads & BIOS+other settings listed so that users could actually see how/when/where these chips clock the best. Now I also mentioned that the sample set from that survey is insignificant, hence drawing any conclusions for the rest of the market is guessing at best/dubious at worst. So yeah it'd be so much better if AMD came out much cleaner than just issuing statements about the fix.
I will again reiterate though that the boost behavior is dependent on a huge number of variables, so it isn't all their fault because they had to keep TDP in check as well as maintain backwards compatibility!