• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Has anyone fried DDR3 ram at < 1.975V

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

magellan

Member
Joined
Jul 20, 2002
According to wikipedia's article on DDR3 RAM:

"In addition, JEDEC states that memory modules must withstand up to 1.975 volts before incurring permanent damage, although they are not required to function correctly at that level."

Aren't RAM manufacturer's supposed to comply w/JEDEC standards?
 
One thing to note is that the JEDEC standard doesn't specify a time frame. That voltage for a short period will not permanently damage ram, but prolonged running of voltages that high will definitely cause degradation without proper cooling, of course.
 
I haven't seen any damaged RAM from latest series running at ~1,975V but many IC are just not scalling above ~1.75V or even lower like some LV Microns that were losing stability above ~1.65V.
New IC from nearly all manufacturers have max safe voltage set at 1.975V. It doesn't mean that memory without proper cooling will survive longer work.
 
Here are 1.35V/1.50V Hynix and Samsung DDR3 chips with 1.80V absolute maximum ratings:

http://www.skhynix.com/inc/pdfDownl...am/Computing_DDR3L_H5TC2G4(8)3CFR(Rev1.0).pdf

http://www.samsung.com/global/business/semiconductor/file/product/ds_k4b2gxx46e_1_35v_rev10-2.pdf

Not that the company that actually built the chip should ever be considered authoritative. :D

I couldn't get a definite answer about how long a chip can be operated at its absolute maximums. Some engineers said years, others said months, but I didn't ask specifically about DRAM.
 
Last edited:
The first generation hypers died consistently at 1.7v. To be fair, the first generation hypers died consistently.

Typically running silicon at its absolute maximum ratings drastically shortens the lifespan of it. Sort of like leaving your car at the redline for all your driving. Yes it's rated to run there, no it's not going to last all that long.
 
The first generation hypers died consistently at 1.7v. To be fair, the first generation hypers died consistently.

Typically running silicon at its absolute maximum ratings drastically shortens the lifespan of it. Sort of like leaving your car at the redline for all your driving. Yes it's rated to run there, no it's not going to last all that long.

These days it seems like if memory lasts a few years it's already obsolete.

I wonder if its possible for them to come up w/a matrix of memory survival rates
at specific voltages?
 
Possible? Sure. You're going to have to buy an awful lot of memory and an awful lot of platforms to make it functional though.
 
Here are 1.35V/1.50V Hynix and Samsung DDR3 chips with 1.80V absolute maximum ratings:

http://www.skhynix.com/inc/pdfDownl...am/Computing_DDR3L_H5TC2G4(8)3CFR(Rev1.0).pdf

http://www.samsung.com/global/business/semiconductor/file/product/ds_k4b2gxx46e_1_35v_rev10-2.pdf

Not that the company that actually built the chip should ever be considered authoritative. :D

I couldn't get a definite answer about how long a chip can be operated at its absolute maximums. Some engineers said years, others said months, but I didn't ask specifically about DRAM.

Is there a newer JEDEC standard for DDR3? Because I had thought all manufacturers were supposed to comply w/JEDEC standards?
 
One of the dual Xeon 5600 LGA 1366 HPC nodes at my work has recently fried out two registered, 8GiB DDR3 modules. This despite the fact they were always run at their specified voltage and being under clocked (because there are too many modules per channel to be run at their rated 1600Mhz speed rating). It's taking FFE to get the vendor to replace them too. What's kinda weird is that they take upwards of three DIMM's per channel, instead of the normal 2 seen in the consumer market.
 
Back