[ttfrog]

I see people putting a crapload of voltage into their ram, and I'm wondering just how safe that is for everyday use. I've been reading that a lot of good BH5 chips take 3.0 + vdimm to run @ high fsb w/ tight timings. I know my IS7 can't do this w/out a volt mod (I'm still debating whether I really want to do it), and therefore I haven't been able to squeeze that much performance out of my memory (I know i can get a lot more). So for now, I'm somewhat satisfied w/ where my memory stands, but if I do decide to volt mod my Is7, exactly how safe is it to run 3.0 vdimm 24/7 or at least all of the day? This is a crucial decision for me because although I overclock, I look for a very high level of stability as well as a long lifetime. Also, would having my vcore @ 1.575 damage the cpu or shorten the lifetime of the cpu? TY FOR THE HELP!!

[shiyan]

Welcome to the forums!

not sure about the RAM voltage, but 1.575v for the CPU is very safe.

would voltmodding and overclocking your RAM a little more really be worth it? increasing RAM performance won't increase your system speed the way increased CPU speed would.

[JKeefe]

As with all overvolting, nothing can be considered 100% "safe". The safest there is is a manufacturer's sanctioned overvolt, such as Corsair recommending 2.75V on my RAM, even though the DDR standard is 2.5V. As such, I would have no problems RMAing my memory if it died at 2.75V.

So, with shiyan saying "1.575v for the CPU is very safe", I wouldn't really call it very safe, I'd call it very safe as far as overvolting goes. With any overvolting, you need to be aware of the risk, and the higher the voltage, the higher the risk.

Obviously there has been documentation of high voltage killing CPUs (SNDS @ >1.7V Vcore). Below that you should be "safe". Below 1.6V you should be "very safe".

I have never heard of anyone killing their DDR from high voltage, even voltage >3.0.

[ttfrog]

hey guys. Thanks for the answer. I guess if i feel venturesome I'll try the volt mod and put the vdimm to 3 volts, which I'm sure would bring sufficient returns with my buffalo 3700 ddr. ty once again

[larva]

The only thing of note I see with running 3V on the ram as I think suspend to ram (S3) mode is a bad idea. My machine tends to be wholely unstable after coming out of S3 suspend and I believe the ram gets hot while it sits with power on it but no airflow. I'm not postitive on this point, but that is the way it appears to me.

Definitely volt mod the IS7, use 400 Ohms on the resistance, but upgrade the power supply first. The TP330 is already likely to limit you on the 12V end and will likley not like voltmodding the memory either, it's hard on the 3.3V rail. These supplies often just blow up under heavy load, and I think the 12V is probably limiting your overclock. 263fsb is nothing for a 2.4c, With 3V on the ram, 1.65V on the cpu, and a good power supply you likely can run 285+ fsb. Some users are as high as 313 or so on the IS7.

[Perre]

Just as a sidenote my OCZ PC3700 Gold leaves me with warranty up to 2.8v :P

[ajrettke]

I've ran RAM at 3.1v for well over 1.5 years (it was crucial PC2100 and it ran at 170 FSB). I would run 3.2v on RAM if it gave me 5 or so mhz more and feel comfortable that it wouldn't blow up. 2.9 and 3.0 is a cake walk.

[flapperhead]

one of the resistorheads in this forum, working on his ee degree did a masterful job of explaining how overvolting leads to component destruction due to electron migration. cant remember thread but if u do a search u might find it....

[JKeefe]

Electro migration always happens. Overvolting components accelerates the process, and it is an endless argument as to how much faster the EM happens, and if it really matters.

[tweaker]

well i guess runnin 3.6volts to my 3500 ocz is overkill..lmao.Anyway its runnin 3.2.2.2 timings(478mhz)

10by238@2380
i ran 3dmark01 at that and yielded 20795(i need to post it)
the other marks r in my sig

later radman