• 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 overclocking kill a GPU?

X

XFreeRollerX

n00b
Joined
Jul 20, 2008
Messages
22
Okay well I've been wondering this for a while now...

I've got two GTX 285's and my first set died on me. They're EVGA FTW editions so they have a slight bump in volts.. THey come stock at 702mhz and I tried to clock em up to 850mhz

Well my OC attempt failed by within a month or so after that, I had to RMA my cards.. Now I've kinda curious... can what I did cause a GPU to just die out like that?
 
If you overclock too much that the heat output is greater than the capacity of the heatsink to cool it then, yes.
 
I fried two ATI cards due to overclocking, the first a Radeon x800 PRO and the second a HD 3850. Both overclocks were well within limits, but overtime they gave out. I figure I just got unlucky, but yes it can happen.
 
Depends. Sometimes things can happen like voltage regulators blowing, caps, etc. But usually the GPU will fry if contact with a heatsink isn't satisfactory. Pushing the ram sometimes with not enough volts can and has damaged cards as well. Stay within your limits.
 
wizang said:
I fried two ATI cards due to overclocking, the first a Radeon x800 PRO and the second a HD 3850. Both overclocks were well within limits, but overtime they gave out. I figure I just got unlucky, but yes it can happen.
Click to expand...

Interesting you say that. Some other members on another forum unanimously agreed that It is impossible for a card to die from raising the GPU clock.


Mind you, I'm no noob with an i7.. I been building and over-clocking systems for years.. but I've never had a card just give up like that... let alone TWO at the same time.

I bought a better PSU after it happened. New GPU and PSU went in at the same time... but I might start OCing again as I like making Crysis my lil woman :D
 
iono if it was due to an oc but my old geforce3 ti200 died after i pushed a 250mhz at the core and around 275memory. it was watercooled too so i don't think heat got it.

as i was working up to that speed i had a bunch of artifacts and system freezes in 3dmark01. not sure what happened but it died anyways

on my 6600gt i haven't had any troubles but i rarely OC it cuz i don't play many games
 
XFreeRollerX said:
Interesting you say that. Some other members on another forum unanimously agreed that It is impossible for a card to die from raising the GPU clock.
Click to expand...

Out of curiosity, what other forum is this?
 
Two at once? Mighta been your rails not working up to snuff on your psu.
 
XFreeRollerX said:
Interesting you say that. Some other members on another forum unanimously agreed that It is impossible for a card to die from raising the GPU clock.


Mind you, I'm no noob with an i7.. I been building and over-clocking systems for years.. but I've never had a card just give up like that... let alone TWO at the same time.

I bought a better PSU after it happened. New GPU and PSU went in at the same time... but I might start OCing again as I like making Crysis my lil woman :D
Click to expand...

I' am knid of curious since when did owning an i7 imply someone to be a newb?

I have heard of a lot of cases involving voltage mods killing cards....I do not see however where ocing could ever kill a card, usually they lock up the system at worst and have to be set to lower clocks.....shorten the life span I could see this happening, running the card hot while ocing....but personally myself the longest I have been between upgrades is a year to two years at the most between new cards,
 
I am dating myself, but I killed a Voodoo 3 3000 while getting carried away. :eek:

I have gotten a bit more conservative in my OC attempts on my 400$+ GPUs as of late. :D
 
Due to my annoyance at purchasing a Matrox Millennium G200 when the nVidia TNT came out a month later (or, rather, I didn't know of it until a month later), I tried overclocking it. It could hardly handle anything above standard, and quickly crashed the computer. After that I got funny dots all over my games. I tried improving things with a fan, but that just made it worse(!)...
 
Usually a short term OC will not kill a card so long as you are not trying for anything extreme. Over time it can kill it, or at the least reduce it's life expectancy.
Overvolting to get your OC is a different matter. Sometime they can take it with good enough cooling, other times they die quickly regardless of how good the cooling is.
 
Modern GPUs its hard to break them, they throttle down when too much heat is detected.

Running high overclocks will tend to lower the lifespan of the card however.
 
AlvinMaker said:
I am dating myself, but I killed a Voodoo 3 3000 while getting carried away. :eek:

I have gotten a bit more conservative in my OC attempts on my 400$+ GPUs as of late. :D
Click to expand...

I killed a Voodoo 3 back in the day overclocking too.:D Had another V3 that was bulletproof and is running in a server to this day, lol. Also ruined a GF4 by overclocking it so it's not limited to one brand. It's not hard to ruin a videocard by overclocking is what I've learned. :p

These days I stick to stock GPU clocks and only overclock my CPU. Overclocking GPU's only gets minimal returns anyway.
 
i'd say you'd typically crash before you managed to do damage from your incredible overclock.

unless you went hog wild with voltage or something.
 
In my experience it is difficult to overclock the core to the point that it will be damaged, especially if you aren't blatantly ignoring warning signs such as artifacting, etc.

Video ram on the other hand... I've seen many cards with video ram that has gone bad due to overclocking and/or overheating. It's always funny when people try to overclock the vram and the chips don't even have any sort of heatsinks on them; or worse, they are "cooled" by the main cooler where they are attached with cheap thermal pads that are 1/16" thick or more and probably insulate more than anything.
 
XFreeRollerX said:
I've got two GTX 285's and my first set died on me.
Click to expand...

Two video cards dieing from overclocks isn't likely. There is some kind of other problem involved. Yeah, video cards can be damaged by overclocks but its rare enough that both of them, no way. Also, almost never do they just "die", you normally get some kind of artifacts. You should probably check your PSU and your PCI-16 slots and make sure they are working.
 
The only thing that I have fried overclocking is a PN5-D Motherboard. Handling overclocked GTX285 and Q9450.
 
OCing is almost guaranteed to shorten the life of the product. Now whether that is shortened from 20 year lifespan to 19 years 300 days, or from 20 years to 20 seconds, depends largely on what components are being stressed and how much stress you put on them.

Heat is one form of stress, and usually the most protected against. Overvolting is a different matter entirely and can quickly destroy a capacitor, transistor, etc. It is entirely possible that both cards failed due to a specific component failing from overvolting on both cards at the same time. Now add stress from overvolting and heat together and that can be a deadly mixture.

All of that being said, I would probably check the power supply first. It sounds more like you pushed a smaller or cheaper power supply farther than it was willing to go, which increased amperage and decreased voltage, which pretty much killed the cards.

I learned a long time ago that for me personally, the amount of reward from OCing did not offset the risk of damage to hardware. All my hardware is moved from my main machine, to my secondary machine, to my wife's machine, and finally to my bench machine as it ages so lifespan is much more important to me than a very slight increase in speed.

Allan
 
well let me ask u somthing if i tryed to push u to far wouldnt u die to?? also u have to remember somtime softwae can crash or shred all it take for software or hardware is one little thing i blew a q9650 trying for 5.8ghz and had to get it rma'd told em it just died but
hard ware is like life one way to do it right unknown number of ways to jack it up
 
i oc'ed my q9650 and it blew out but it was actualy the bottom and not the top one thing i do is extensive research on everything i buy aqure mess with etc... less failures read others failure dont make theres agian let others f it up and u learn ..............cheaper that way. i can run my q9650 @ 4.9 stable if i wanted to but just cause i can dont mean i should heck i rn a pentium d 3.4 # 10.7 once just for the heck of it lasted 2weeks nevergot above 30c
 
well let me ask u somthing if i tryed to push u to far wouldnt u die to?? also u have to remember somtime softwae can crash or shred all it take for software or hardware is one little thing i blew a q9650 trying for 5.8ghz and had to get it rma'd told em it just died but
hard ware is like life one way to do it right unknown number of ways to jack it up
Click to expand...


i oc'ed my q9650 and it blew out but it was actualy the bottom and not the top one thing i do is extensive research on everything i buy aqure mess with etc... less failures read others failure dont make theres agian let others f it up and u learn ..............cheaper that way. i can run my q9650 @ 4.9 stable if i wanted to but just cause i can dont mean i should heck i rn a pentium d 3.4 # 10.7 once just for the heck of it lasted 2weeks nevergot above 30c
Click to expand...



Alona... there was so much fail it those two posts I am embarrassed for you. :eek:
 
meh it seems to be a 50/50 here, some people have a clue and others apear to be completely clueless and spew garbage as if it is the divine word of god and have obviously just recently like yesterday started to play' around with computers.....killing and frying components evidently seems to them to be a pedigree to being some sort of tech wiz.....killing 2 280....I have to ask myself how in the hell did you manage to do this...did you throw water on them...wtf? haha. Consider yourself lucky when that 99% of the modern hardware is basically fry proof but unfortunately not idiot proof, unless you are total nub and try crazy shit with it.....(tip: unless you like negative feedback you should keep really stupid shit you do or try to yourself) if you could put a fart can on it you probably would.....how many fast and furious honda civic owner we got on her btw?
 
I've never seen CPUs or GPUs fail from overclocking (because you simply can't tell). I've had a few that, over time, lose their ability to clock as high as when new, but never fail in a way that would convince me that it was an overclocking issue. Barring heat and way too many volts, its impossible to say whether a component died from overclocking.

Its like those cold medicines that claim to reduce the duration of a cold. How do you know you would have been sick for a longer amount of time? Same goes for processors. How can you determine that your processor would have lasted longer? You can have one card overclocked and another stock speeds, and again, barring temperature issues or overvolting, either one might fail first.
 
alona said:
well let me ask u somthing if i tryed to push u to far wouldnt u die to?? also u have to remember somtime softwae can crash or shred all it take for software or hardware is one little thing i blew a q9650 trying for 5.8ghz and had to get it rma'd told em it just died but
hard ware is like life one way to do it right unknown number of ways to jack it up
Click to expand...

alona said:
i oc'ed my q9650 and it blew out but it was actualy the bottom and not the top one thing i do is extensive research on everything i buy aqure mess with etc... less failures read others failure dont make theres agian let others f it up and u learn ..............cheaper that way. i can run my q9650 @ 4.9 stable if i wanted to but just cause i can dont mean i should heck i rn a pentium d 3.4 # 10.7 once just for the heck of it lasted 2weeks nevergot above 30c
Click to expand...

w...o...w........... I'm speechless.
 
Mega666 said:
Amps kill, not MHz.
Click to expand...

High temperature kills, higher power draw causes higher higher temperatures, and higher clocks require higher power draw, just look at the 5970, a card with 2x 5870 worth of hardware installed but only clocked at 5850 speeds to stay under the 300W power limit.

Most cards are protected against damage from heat due to automated systems that kick in and throttle the speed of the card at high temps. You could probably still kill cards with overvolting if you edit the BIOS to have a stupidly high stock voltage or a manufacturer was silly enough not to limit the digital voltage regulators on the card to protect the sensitive components.
 
Dethred said:
I've never seen CPUs or GPUs fail from overclocking (because you simply can't tell). I've had a few that, over time, lose their ability to clock as high as when new, but never fail in a way that would convince me that it was an overclocking issue. Barring heat and way too many volts, its impossible to say whether a component died from overclocking.

Its like those cold medicines that claim to reduce the duration of a cold. How do you know you would have been sick for a longer amount of time? Same goes for processors. How can you determine that your processor would have lasted longer? You can have one card overclocked and another stock speeds, and again, barring temperature issues or overvolting, either one might fail first.
Click to expand...

Man I wish I had the pics from my nV Geforce II Ultra the black marks on the PCB would have shown you the way, my friend and I strapped an 80mm fan on that beast. Back then we were trying to get high scores in 3dmark and as it turns out that card had a limit. Of course it’s somewhat blurry as I believe we were a bit drunk and didn’t stop once we saw artifacts..
 
I can ALMOST still smell the smolder that came off my Voodoo 3 (3500? I can't recall) from overvolting.

Yes, I got power hungry...and paid the price. :D
 
PrincessFrosty said:
High temperature kills, higher power draw causes higher higher temperatures, and higher clocks require higher power draw, just look at the 5970, a card with 2x 5870 worth of hardware installed but only clocked at 5850 speeds to stay under the 300W power limit.

Most cards are protected against damage from heat due to automated systems that kick in and throttle the speed of the card at high temps. You could probably still kill cards with overvolting if you edit the BIOS to have a stupidly high stock voltage or a manufacturer was silly enough not to limit the digital voltage regulators on the card to protect the sensitive components.
Click to expand...

Your right but your wrong at the same time...sure heat can shorten the lifespane of components... seriously though anyone who is not a fuctard is going to take cooling into consideration beforehand if they intend on ocing thier hardware....atleast I would thinks so....the decrease of EOL is a given... but again anyone who is into this hobby is most likely going to replace thier cpu/gpu/mobo/ram etc.....far before the dereased lifespan would have any real impact....to me here is the deal though....99.9% MODERN (aka not a vodoo gpu) cpu/gpu have some sort of thermal shutdown built into their bios or/and hardwired to prevent and outright meltdown....you cannot fry a cpu today and make it explode like you could a AMD Athlon XP which you could melt merely by setting the clocks too high.....naw today you actually have to get very inventive and find ways around these safegaurds.....one of the most popular way is volting/aka volt mod and it has the ability to render a cpu/gpu dead in a heart beat SO YES AMPS used stupidly =death

Little experiment I did with an old XP I had, I took off the hsf and within 30 seconds it began to literally smolder....blackish soot and smoke and all.....did this with a e8400 also....the e8400 simply shut down and remained undamaged.........pretty much confirms the idiot proofing.
 
Last edited:
Tabasco69 said:
Your right but your wrong at the same time...sure heat can shorten the lifespane of components... seriously though anyone who is not a fuctard is going to take cooling into consideration beforehand if they intend on ocing thier hardware....atleast I would thinks so....the decrease of EOL is a given... but again anyone who is into this hobby is most likely going to replace thier cpu/gpu/mobo/ram etc.....far before the dereased lifespan would have any real impact....to me here is the deal though....99.9% MODERN (aka not a vodoo gpu) cpu/gpu have thermal shutdown built into their bios or/and hardwired to prevent and outright meltdown....you cannot fry a cpu today and make it explode like you could a AMD Athlon XP which you could melt merely by setting the clocks too high.....naw today you actually have to get very inventive and find ways around these safegaurds.....one of the most popular way is volting/aka volt mod and it has the ability to render a cpu/gpu dead in a heart beat SO YES AMPS used stupidly =death
Click to expand...

Funny you should mention the Athlon XP...I had one of those too...but wasn't so careless :cool:
 
Of course you can kill any chip overclocked. There is a reason why there at a certain stock speed in the first place. If you were to do a deep serch in this forum you will notice the ones with the most PC part problems are the dudes that are OC'd to the tits! I have never lost a single motherboard, vid card, hard drive or stick of ram since I built my very first PC back in 1998. Guys that OC the crap out of there parts are going through motherboards and vid cards and cpu's and power supplies all the time. I just sit back and laugh reading it all. I mean think about it, any money they thought they saved by OC to much faster speeds that lost by re-buying blown parts LOL
 
maleficarus said:
Of course you can kill any chip overclocked. There is a reason why there at a certain stock speed in the first place. If you were to do a deep serch in this forum you will notice the ones with the most PC part problems are the dudes that are OC'd to the tits! I have never lost a single motherboard, vid card, hard drive or stick of ram since I built my very first PC back in 1998. Guys that OC the crap out of there parts are going through motherboards and vid cards and cpu's and power supplies all the time. I just sit back and laugh reading it all. I mean think about it, any money they thought they saved by OC to much faster speeds that lost by re-buying blown parts LOL
Click to expand...

Ehh. I've had video cards that ran stock speeds 24/7 end up go bad on me. Conversely I've also had chips OC'ed that are still working to this day. Today OC-ing is covered by most product manufacture's warranty so you might as well do it, assuming you have the cooling to spare.
 
Last edited:
Haven't killed anything overclocking. Remember that you're working with a precision electronic component that was designed to work within a set a parameters. Even if the same i7 wafer is used for the 920,945,and 960, those chips have been hand picked for use with a higher clock. Same with anything else. Watch for warning signs.
 
You must log in or register to reply here.
Back
Top