Should I keep 1.75V for 24/7 DDR3 OC?

[damric]

Since my board won't let me reach the 2133 CL9 that my 2x4GB Ripjaws are rated at, I've tightened the timings down at 1800 to CL7, but it took a bit of juice to do it.

 

[Outback Bronze]

I voted yes. I cant see why not. Ive never seen memory blow up.

 

[MxPhenom 216]

it'll slowly kill the ram if you run it at that voltage.

 

[Daimus]

As for the slow death of memory, Gskill has a lifetime warranty on the modules. If memory is not overheat and stable, why not?
I voted "yes."

 

[erocker]

Perhaps you should try upping the CPU/NB voltage and lowering the RAM voltage. See if it works.

 

[n-ster]

Perhaps you should try upping the CPU/NB voltage and lowering the RAM voltage. See if it works.

golden advice

 

[The Von Matrices]

I ran my old Core 2 Quad with 1.95V DDR3 for three years and it did no harm, so I vote yes. I'd suspect the memory controller rather than the memory would be the first thing to get damaged.

 

[BlackOmega]

I ran my old Core 2 Quad with 1.95V DDR3 for three years and it did no harm, so I vote yes. I'd suspect the memory controller rather than the memory would be the first thing to get damaged.

Bingo. I voted no because of this exact reason. The memory controller is only rated to run RAM @ 1.65v, anything higher than that will damage the IMC. Now mind you, this isn't going to happen right away, it'll slowly degrade over time. So in a matter of a few months you'll be like this wondering why your rig keeps crashing.

 

[erocker]

Core 2's could handle high voltage DDR3, so can AMD, though I'm not positive on bulldozer. I actually still have a set of 1.8v DDR3 laying around that I used on my 1100T. I just don't know if the RAM itself can handle 1.75v for very long, though I'm pretty sure the IMC can take it.

 

[claylomax]

I say crank it up a notch. I've run mine at 1.8v with no issues.

 

[Daimus]

If I am not mistaken, Damric has MSI 890fx GD65. The motherboard should be able to work with the memory at a high voltage, although it is not stated in the description on the MSI official site.
For example, for my mobo chipset 890fx recommended memory voltage to 1.9v.
http://dlcdnet.asus.com/pub/ASUS/mb/socketAM3/Crosshair_IV_Formula/C4F_Memory_QVL20100726.pdf
I doubt that the work at high voltage memory can damage the AMD FX IMC. But it's better to listen to Eroker's good advice .

 

[Woomack]

I doubt that you can damage memory while running @ 1.75V as most IC on market are designed for ~1.65V +/- 0.1V so 1.75V is about max. New IC have a bit lower voltage but still you can find memory on the same chips in 1.35-1.65V sticks. Other thing is that most new IC are not scalling well above 1.65V and some don't like 1.6V+ so maybe you won't kill them but you can lose stability.
Elpida, Micron, Qimonda and some other IC are running fine till 1.8V. PSC, Hynix and Samsung like lower voltages like 1.35-1.65V ( even that PSC are sometimes labeled as 1.55-1.75 then most are almost not scalling above 1.7V on air cooling ) ... That's just my experience and I can be wrong about some IC.
Damaging IMC by too high memory voltage is Intel's theory ( remember that they weren't even sure if it's true when they said that but you can see it in specification for Intel 6 series chipsets ). I haven't seen anyone damaging IMC on P67/Z68 board running memory @1.7V+ and I was running some sticks @1.8V for long tests and ~1.7V 24/7.
If sticks won't overheat then I see no problem. There is almost no difference in performance on FX if you run memory @CL7 or CL9 so I see no point to heat up memory with 1.75V. Better try to set higher CPU-NB that is easier to stabilize than memory@2133 and gives better results.