VDDQ is not RAM voltage though, and that is my point. Your contention is based upon a flawed premise.

Ok, Max. I don't know if Vddq is the same as Vdimm or not, but just so that you see I didn't pull this one out of a hat, here's just one example I found after a quick search:http://i4memory.com/...y-ratios-23430/Not saying it proves anything, just that Vddq absolute maximums is what everyone has used for Vdimm for yearsIf you ask in any overclockers forum out there no one will tell you 1.8V is not fine for Nehalem, and dozens of them have been running 1.7+ for 4 years or more

We're going to find out, because a whole lot of us are using 1.65V...and a number of memory vendors are saying in their forums that 1.65 is fine. Mushkin and Patriot have said we are good to go.K

Of course they're saying it's ok. Anything else would make them liable for a world of hurt. They're likely pinning their hopes on a few factors:1) it is not their product which will fail, but rather Intel's product2) the majority of users of such exotic RAM are likely to run their CPUs out-of-spec in which case they can blame the user for the failure3) most power-users upgrade too quickly to experience total failure

Ok, Max. I don't know if Vddq is the same as Vdimm or not, but just so that you see I didn't pull this one out of a hat, here's just one example I found after a quick search:http://i4memory.com/...y-ratios-23430/Not saying it proves anything, just that Vddq absolute maximums is what everyone has used for Vdimm for yearsIf you ask in any overclockers forum out there no one will tell you 1.8V is not fine for Nehalem, and dozens of them have been running 1.7+ for 4 years or more

People can assign whatever meaning to whatever they like, it does not make it correct, however. One thing to keep in mind here is that any resultant failure will not be immediate, it will take months or even years to manifest. Re: VDDQ vs. VDIMM: VDDQ is a value defined by Intel that describes the voltage for the memory I/O controller integrated into their processors which feature this component (i.e. Core i series CPUs). This is voltage applied to certain circuits within the CPU itself. VDIMM is the value defined by motherboard manufacturers as the voltage supplied to the RAM module(s).

People can assign whatever meaning to whatever they like, it does not make it correct, however. One thing to keep in mind here is that any resultant failure will not be immediate, it will take months or even years to manifest. Re: VDDQ vs. VDIMM: VDDQ is a value defined by Intel that describes the voltage for the memory I/O controller integrated into their processors which feature this component (i.e. Core i series CPUs). This is voltage applied to certain circuits within the CPU itself. VDIMM is the value defined by motherboard manufacturers as the voltage supplied to the RAM module(s).

Years to manifest? Then who cares? I'm kidding. I just never keep anything for years with FSX. Maybe 18 months, (my first i7), but that's it...I ran my q6600 @ 3.6, my q9650 @ 4.0 and my i7-860 @ 4.2. That probably reduced the life of all of them, but I didn't keep them long enough to see...

People can assign whatever meaning to whatever they like, it does not make it correct, however. One thing to keep in mind here is that any resultant failure will not be immediate, it will take months or even years to manifest. Re: VDDQ vs. VDIMM: VDDQ is a value defined by Intel that describes the voltage for the memory I/O controller integrated into their processors which feature this component (i.e. Core i series CPUs). This is voltage applied to certain circuits within the CPU itself. VDIMM is the value defined by motherboard manufacturers as the voltage supplied to the RAM module(s).

And if Vdimm is simply the voltage in the RAM modules, what does that have to do with Intel (CPU or chipset) and why would they even bother setting a maximum?Another question, have you ever seen that Vddq voltage in any Nehalem system BIOS? There's a Vcore (Vcc), a Vtt, A Vpll... everything in that table of absolute maximums but no Vddq :Thinking:

And if Vdimm is simply the voltage in the RAM modules, what does that have to do with Intel (CPU or chipset) and why would they even bother setting a maximum?Another question, have you ever seen that Vddq voltage in any Nehalem system BIOS? There's a Vcore (Vcc), a Vtt, A Vpll... everything in that table of absolute maximums but no Vddq :Thinking:

Reading Intel's datasheet it is clear that VDDQ is a voltage domain for the processor, not some external device (such as RAM). I can't find it in the data sheet (neither do I have time to examine all of them to find this information) but if you simply reference the numerous launch reviews that all state this, and the stickers motherboard manufacturers place on all x58 boards, it is quite clear that 1.65V is the recommended maximum. Intel doesn't intentionally give out incorrect information to customers nor the press, so you can rest assured that if they had, it would have been corrected. Since it has not, it is safe to assume this information is correct. Here's a pre-launch news piece on the matter, which reflects this fact. It is stated in the piece that the journalist confirmed this limitation with Asus.

Reading Intel's datasheet it is clear that VDDQ is a voltage domain for the processor, not some external device (such as RAM). I can't find it in the data sheet (neither do I have time to examine all of them to find this information) but if you simply reference the numerous launch reviews that all state this, and the stickers motherboard manufacturers place on all x58 boards, it is quite clear that 1.65V is the recommended maximum. Intel doesn't intentionally give out incorrect information to customers nor the press, so you can rest assured that if they had, it would have been corrected. Since it has not, it is safe to assume this information is correct. Here's a pre-launch news piece on the matter, which reflects this fact. It is stated in the piece that the journalist confirmed this limitation with Asus.

That recommandation keeping voltage at 1.65 was a thumb of rule for the X58 chípset. The truth is that with a X58 chip your on the safe side if you keep DRAM bus voltage within 0.5V diff compared to the QPI/DRAM (uncore) voltage. If your memory sticks would handle 1.85V that would not be any problem as long as you set your QPI/DRAM voltage at min 1.35V. As far as I remember the max QPI/DRAM voltage is 1.375V (info provided by Intel).Asus, Intel and other HW companies will provide us with recommendations/thumb of rules aimed at the noob consumer market. To obtain the real facts you need to use other sources.

That recommandation keeping voltage at 1.65 was a thumb of rule for the X58 chípset. The truth is that with a X58 chip your on the safe side if you keep DRAM bus voltage within 0.5V diff compared to the QPI/DRAM (uncore) voltage. If your memory sticks would handle 1.85V that would not be any problem as long as you set your QPI/DRAM voltage at min 1.35V. As far as I remember the max QPI/DRAM voltage is 1.375V (info provided by Intel).Asus, Intel and other HW companies will provide us with recommendations/thumb of rules aimed at the noob consumer market. To obtain the real facts you need to use other sources.

You're entitled to believe whatever you like but to blatantly ignore manufacturer warnings is walking a thin, dangerous line.