Wouldn't be affected by
what? The power dissipation of the memory? That was my whole point!
I'm not entirely sure what you're talking about, with your comment about how it "can't be measured separately", but
I was referring to the article, which quoted TDP numbers from AMD (click through the various slides - they have one titled "AMD AM4 Chipset Efficiency") and made a speculation which I quoted. I won't repeat that, but check my post & the article.
Indeed,
wikipedia indicates that AMD 970 was built on 65 nm. I checked a page on AMD's site for it, but didn't see any datasheet links. Didn't really care to look further, because the only thing that matters now is what process the new chipset uses. But then I didn't care enough to look for that, if they even announced it (the above wiki page doesn't currently know it).
Xajel :
Sorry bro but you're wrong, heat generated does not equal to power consumption, the heat is an un wanted by product duo to several reasons sources,
Sorry bro, but TJ knows his stuff. Power consumption = heat output. Sure, some miniscule amount of energy is dissipated in other forms, but very nearly all of it is converted to heat. Most of us know that. I suggest you do some more reading on the subject.
The one distinction that
can be drawn is that TDP represents a typical
peak level of power consumption. So, you could have two chips with the same TDP, but one might have lower
average power consumption. That's more concern for tablets & laptops, since battery life is a significant issue for those applications. But, for most of us desktop users, AC power is cheap enough that we're mainly concerned with cooling (which must be designed around the sustained peak output). Hence, TDP gets most of the attention among us, with idle power being only a secondary or tertiary concern.