• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Can a GPU's memory “die” through overclocking?

Coldblackice

[H]ard|Gawd
Joined
Aug 14, 2010
Messages
1,156
I hear people comment about GPU memory/ram "dying" from an overclock. Is this possible? Is there an objective way to tell, besides circumstantial blue screens?
 
It all depends on how far you're pushing it and how long your running it for.

Try to avoid artifacts.
 
Yes, it can... on older cards it would artifact (GDDR3/below), but with GDDR5's error correction it won't, rather it will just degrade performance unless you push it extremely far, where it can just overstress the memory controller and/or VRAM and the card will have issues (i.e. not POST, have artifacts, cause crashes, or any other number of things).
 
Yes, it can... on older cards it would artifact (GDDR3/below), but with GDDR5's error correction it won't, rather it will just degrade performance unless you push it extremely far, where it can just overstress the memory controller and/or VRAM and the card will have issues (i.e. not POST, have artifacts, cause crashes, or any other number of things).
gddr5 will most certainly artifact and will crash if clocked too high. on my 670 there was a spot where performance would degrade just slightly and any higher would artifact and then crash. on my 660ti it seems to level off at a certain spot then any higher will artifact and crash. by crash I mean what ever I am testing will lock up or just stop if I let it keep running after seeing artifacts.
 
yea you bet it can, you can blow a lot of stuff if you overvolt it too much
then you have to take into account the extra heat
 
My card wouldn't show artifacts but when it did mess up the screen would just turn solid gray or some full screen effect like that would happen and I'd have to restart. Never artifacted though.
 
gddr5 will most certainly artifact and will crash if clocked too high

Yes, thanks for repeating what I already said... bold emphasis mine:

Yes, it can... on older cards it would artifact (GDDR3/below), but with GDDR5's error correction it won't, rather it will just degrade performance unless you push it extremely far, where it can just overstress the memory controller and/or VRAM and the card will have issues (i.e. not POST, have artifacts, cause crashes, or any other number of things).

You sound like you're trying to make a correction when you're just paraphrasing what was already said by me. :confused:
 
Yes it has happened many times according to forum reports. I've been left wondering if the vram voltage is adjusted slightly with speed automatically. Because it seems unnatural to simply change freq and then induce persistent failure within months. Usually the culprit is voltage, since power/heat rises with the square of voltage, but only linearly with frequency.
 
I hear people comment about GPU memory/ram "dying" from an overclock. Is this possible? Is there an objective way to tell, besides circumstantial blue screens?
By using the word "dying," I assume you're asking about permanent damage. In that case it's very difficult to kill a component, even something more sensitive like vRAM, simply with overclocking. Overvolting is really what damages hardware. That said, you can shorten the lifespan of vRAM by overclocking if it's not adequately cooled. The increased speeds and heat accelerate electro migration.
 
Back
Top