damric :
I knew right away that you would quote that article.That's not Intel. That's MJ. If you do some google research, all of these false rumors around the internet circle around and around and eventually end up with MJ and his mysterious and absolutely unconfirmed "insider source" of which he won't show me the "email correspondence". I've dogged him to death over this for months. He's got nothing.
We had similar misinformation spread about the several previous Intel CPU architectures, people saying that they would break if we used 1.65v DDR3. Intel gave all of these CPUs including Skylake DDR3 memory dividers that go beyond 3000MT/s, of which none of them would work if indeed we were not supposed to add some voltage, which is still within the specifications of JEDEC.
Hmm, got it. I've found it quite strange that manufacturers make DDR3 boards because of the returns and PR nightmare they would be risking if the info in the article is legit. Now I'm thinking very very unpleasant thoughts about how that article might have come to be...
Regardless, honestly, would you personally "risk" it in your Skylake rig if you had one? I wouldn't. Hence, I cant bring myself to recommend that and would rather try and convince people to forget about and sell their DDR3, which, as you may imagine, is, more often than not, a hefty proposition.
All that being said, this merits further inquiring.