For anybody making claims about the "rest of the system" when looking at power draws for the 6700, go look at the rest of the review more carefully. The 6700 platform actually has *lower* idle power draw than any of the other systems (igp 6700 beats igp 4770K by about 4 watts actually). Using the 6700 (igp mode) in a CPU heavy task where there's nothing else going on makes you wonder what else is drawing all that power.. I doubt it's the RAM using that much juice.
Here's the TR graphs of the idle vs. load powers:
Look at the delta for the IGP 6700:
Idle: 23 watts (best score on the chart BTW)
Load: 103 watts (higher than the 4770K with a discrete GPU by 1 watt now, and 16 watts higher than the IGP 4770K under load)
Delta: 80 watts
But But BUt... you say that there are other things than the CPU that take up power.
Well that's technically true, but let's do some overly favorable estimation in favor of AMD here:
1. Assume that the 6700 uses literally zero power at idle. We're talking idle power draw that's better than your smartphone here.
2. Assume the 6700 gets right up to that 65 watt power draw when railed with x264.
3. So: 80 watt observed power delta - 65 watts used by the CPU = 15 watts.
4. So, when the CPU is being railed, the rest of the platform is using 15 + 23 = 38 watts to do... what exactly?
-- It's MEMORY!! Well, if it is, then the RAM must be using just about zero power at idle just like our miraculous CPU to show a delta that big. Once again, the other platforms are using RAM too, why is the delta so big for the 6700?
-- It's Disk!! Really? The other platforms used the same storage you know. Once again, why is the delta that much bigger than the max TDP for a clearly CPU-bound task?
5. As soon as you start raising the idle power of the CPU, the delta starts to get even harder to explain than what we have above. Question: How many systems other than the 6700 and 6800K have power deltas that are actually larger than the TDP of the corresponding chip in those graphs?