Should I run memory at 2.4V?????

[memorymaster]

Now, I am pretty confused about if I should run memory at 2.4V, even though alot of the main memory companies have high performance ram rated for 2.4V.

From EVGA's own press release:

http://www.evga.com/articles/350.asp

"Important message regarding memory and the EVGA 680i motherboard

NVIDIA has investigated end user reports of high performance DIMM failures on the NVIDIA nForce 680i SLI-based platforms. During this process we have been in close contact with DIMM manufacturers and the DRAM manufacturers they rely on to understand the failure scenario. By working with our community, we believe that the observed failure is a breakdown of the silicon in the DRAM caused by the prolonged application of 2.4V on the voltage rails of the DIMMs.

NVIDIA own internal testing has observed this failure on multiple motherboards using different chipsets (both NVIDIA and non-NVIDIA chipsets). This issue is not directly related to motherboards using the NVIDIA nForce 680i SLI MCP or other chipsets.

If you are using this type of memory and are experiencing this issue, NVIDIA recommends contacting your memory manufacturer or system manufacturer for additional information and warranty information."

So should I not try to overclock with 2.4V, or should I go with what the memory companies say by trying the high end memory at 2.4V and RMAing the memory every other week?

Anybody else have a kit running at 2.4V?

 

[subtlewordplay]

It depends on what kind of gain you would be getting. You definitely would need active cooling on the ram. It would be well-advised to purchase ram with 2.4 EVP warranty. There was a big issue with EVGA 680i and high voltage recently. I'm not sure if it's a particular board issue or chipset issue. But proceed with caution.

 

[Stewartwi]

If memory serves me correctly EVGA 680I boards were having issues with ram running above 2.1V

 

[Track]

If memory serves me correctly EVGA 680I boards were having issues with ram running above 2.1V

680i SLI's have many issues. All will be fixed.

To the OP:
You should run ur RAM as high as u need for the overclock u want.
Just remmember that anything above 2.4v leaves u without a warranty.

 

[Mondoman]

Well, if you are running DDR RAM, 2.4V is undervolting it, so I'd increase to at least the standard 2.5V in that case.

I only know of one manufacturer warranty on DDR2 that runs up to 2.4V operation, and I don't know of any manufacturer rating their RAM at 2.4V (2.2V is the highest I've seen). Since 2.4V is a 33% increase over the standard 1.8V that all these modules are designed for, and since electromigration is a real issue that cannot be "solved" by cooling, it seems pretty silly to me to run at such high voltages. YMMV.

 

[Stewartwi]

If memory serves me correctly EVGA 680I boards were having issues with ram running above 2.1V

680i SLI's have many issues. All will be fixed.

To the OP:
You should run ur RAM as high as u need for the overclock u want.
Just remmember that anything above 2.4v leaves u without a warranty.

No doubt not trying to diss the board at all just informing the OP. I have an EVGA 680I and I love my board never had any problems with it at all. Just trying to be helpful.

 

[Stewartwi]

I run these myself. Haven't had any problems with them on my EVGA granted I have them running at 2.1v.

 

[memorymaster]

I only know of one manufacturer warranty on DDR2 that runs up to 2.4V operation, and I don't know of any manufacturer rating their RAM at 2.4V (2.2V is the highest I've seen).

I know that G.Skill, Corsair, OCZ, G.Skill are all at 2.4V, with Kingston at 2.35V.....

But electromigration seems like a big "problem" for the most recent memory modules, but isn't electromigration heavily depended on temperature (as well as other factors), so any kind of cooling would help, right?

Or is the fan just a gimmick? 😀

Thanks for all the input you guys have given me!

 

[subtlewordplay]

Essentially "electromigration" is just another term for the buildup of heat. Any kind of substantial overclocking will generally heat up the ram. A fan would only be beneficiary to the cause.

You might even want to try cooling the chipset.

But generally the lower the voltage you can go to get them running the better.

 

[Mondoman]

...
I know that G.Skill, Corsair, OCZ, G.Skill are all at 2.4V, with Kingston at 2.35V.....

Yep, it looks like they've had to go that high for their very "fastest" DIMMs. I guess they figure the price premium will cover the warranty return costs. Must be enough customers with more money than common sense...

...But electromigration seems like a big "problem" for the most recent memory modules

Only on the ones that are being massively overvolted, like those above. Here's an analogy: you can take a standard car engine and bolt on a turbo, and sure enough, it'll produce a lot more horsepower. However, the more boost you use, the sooner the engine blows, because it wasn't designed for such pressures. Me, I'd rather buy a car that was designed to use a turbo.

.., but isn't electromigration heavily depended on temperature (as well as other factors)

Mainly voltage, then temp.

.. so any kind of cooling would help, right?

Not really. Any kind of cooling *that results in lower temps on the internal traces/junctions" will help. If you have an oven and you point a fan at the outside, that's not going to cool off the inside of the oven very much. However, if you run a heat pipe from the inside of the oven to a heatsink on the outside, that'll work much better. That's why DIMMs designed from the ground up for enhanced heat transfer through the chip contacts (e.g. Corsair's Dominator) can make some difference, but just pointing a fan at a standard DIMM won't be nearly as effective.

 

[memorymaster]

Great explanation of the temperatures... and now it makes so much sense to me. Of course the most heat created will be at the back of the chips, where the FBGA interconnects are...

Guess I'm only gonna be going up to 2.3V with my memory.

 

[Kari]

.., but isn't electromigration heavily depended on temperature (as well as other factors)

Mainly voltage, then temp.

the magic word is current density, but yes, with higher voltage there will be greater current going through.

 

[memorymaster]

Kari,

You and Mondoman seem pretty knowledgeable about electromigration, while I'm not very well educated in the subject.

So, if there is a PCB like corsair's dominator, then outside cooling of the module will help reduce electromigration since the PCB helps extract heat more?

Then is it possible to raise the voltage up to 2.4V with extreme cooling conditions, or does the voltage fluctuate too much on that high of a setting to where the chips will fry regardless?

 

[XrayMan]

I have the Corsair 9136C5D 1142mhz. I have it running at the default voltage of 2.1v on my Evga 680i. I have no problems at all. Also the bios is set to Sli-Memory mode.

 

[Mondoman]

...
So, if there is a PCB like corsair's dominator, then outside cooling of the module will help reduce electromigration since the PCB helps extract heat more?

If the outside cooling increases the heat transfer. I suspect (but don't know) that it won't make much difference due to the majority of the heat being transferred through the conductive leads into the MB.

...Then is it possible to raise the voltage up to 2.4V with extreme cooling conditions, or does the voltage fluctuate too much on that high of a setting to where the chips will fry regardless?

Voltage fluctuations shouldn't be the issue. IMHO, it's silly to run a module at a voltage so much higher than what it was designed for. However, it won't hurt anything besides (possibly) the module itself, and Corsair says they'll warranty it under those conditions, so all you have to lose is the high purchase price of the module (and perhaps some time waiting for it to be replaced if it does fail).

 

[Stewartwi]

I have the Corsair 9136C5D 1142mhz. I have it running at the default voltage of 2.4v on my Evga 680i. I have no problems at all. Also the bios is set to Sli-Memory mode.

That's the ram I'm running and 2.4v is not the default for that ram. It's 2.1v

 

[XrayMan]

That's the ram I'm running and 2.4v is not the default for that ram. It's 2.1v

Your right. It is 2.1v. Sorry, but it was an honest mistake. I have edited that other reply of mine to reflect the correct 2.1v. I am now going to go into my bios to make sure its 2.1v. I don't know why I was thinking 2.4v. 😳

 

[niz]

Essentially "electromigration" is just another term for the buildup of heat.

No it really isn't. Its the transport of material caused by the gradual movement of the ions in a conductor due to the momentum transfer between conducting electrons and diffusing metal atoms.

In other words, Parts of your expensive device are being permanently moved/damaged by the excessive voltage degrading the transistor junctions themselves. This is much more than just heat.

And to the original poster... I'd suggest contacting Corsair directly and getting an update (in writing, e.g. as an email) confirming whether they still confirm OK to run the RAM at 2.4v.

Also don't forget that there will always be some small degree of error in the voltage regulator, in the actual voltage the ram gets is not what you set by some small amount. If the error is in a positive direction you will be actually applying a higher voltage than the value you've set anyway.

 

[XrayMan]

Hmm, voltage was at 1.85v. I upped it to 2.1v. Hope all goes well.

 

[Stewartwi]

Hmm, voltage was at 1.85v. I upped it to 2.1v. Hope all goes well.

Aye shouldn't be an problem. 🙂 Just don't want to see you mess up that ram by over volting it.

 

[XrayMan]

No. I don't plan on over volting it. I think I will just play it safe and stay with what it supposed to be at. I can sleep better at night that way. :lol:

 

[Crashman]

Most RAM will get very little stability gain for overclocking by increasing from 2.2V to 2.4V. Some will even lose stability, due to added heat.