GPU Lifespan ( Mining )

[Exaco]

Hey guys, how much of lifespan decrease i can expect if i leave my Zotac 980Ti at fixed 75c temp limit for 24/7 ?
Its running at 1455Mhz Core and 3304Mhz Mem clock.

[DrMacintosh]

How about you just don't mine? 

 

Lifespan will be very short and the returns with that GPU probably will be lower than the power bill. 

[Exaco]

How much shorter compared to gaming lets say approx ? I can make around 6$ profit per day ( 6.5$ in total ), and power cost per day is maximum 50cents.

I have found this answer on bitcoin talk, so propably thats a lottery, i am also afraid that it can void the warranty ( just asked the shop i've bought from, waiting for reply ), otherwise i can earn 300$ in few months, so that would be the gpu's price, it even dropped to ~200$ now so i basically could earn for a new card in 1 and a half month. But i wonder how really bad it is to keep it running on 75c constant.
 

Quote

I've heard of GPUs mining in the 70Cs that have failed after a couple of months, and others that have run in the 90Cs for years. Typically keeping them cooler will always be better, but also remember that temperature variance will do a lot more harm than you think. Running at GPU at 80C nonstop for months at a time is better than running a GPU up to 70C -> down to 30C -> up to 70 -> down to 40 -> etc.

 

[faziten]

6 hours ago, Exaco said:

Hey guys, how much of lifespan decrease i can expect if i leave my Zotac 980Ti at fixed 75c temp limit for 24/7 ?
Its running at 1455Mhz Core and 3304Mhz Mem clock.

which zotac? 

Mining wrecks VRM, and only the gpu if you apply too much voltage. Of course you can forget about mining with gpu's overclocked 24/7 without side effects on the long therm. Remember you are running out of spec. Also remember memory overclock does not affect mining whatsoever. So drop it to stock or flash a custom bios with underclocked VRAM for mining only, so you can redirect the resources to GPU only. 

 

Life span, is given by temperature, voltage (highly suggested to under volt) and PSU quality. 

 

Don't do it with a 980ti not because it might die, just because it's nonsense the money spent vs money earned. 

[Vellinious]

Running 24/7?  It'd be a miracle if it made it 6 months.  A constant 75c is going to eat that thing alive.

[rn8686]

Although I can make a decent profit with my 1070 (Upwards or $6 AUD/day with power costs included) and am very comfortable with temps (43c ATM), that will kill your GPU with in a few years and is also a very power hungry card canceling out alot more of your profit. You can still make something, just be aware it wont live for long like that. 

[AgusAlexander]

4 hours ago, Vellinious said:

Running 24/7?  It'd be a miracle if it made it 6 months.  A constant 75c is going to eat that thing alive.

74 too high for a GPU? Those are normal temps are you saying this because of the 24/7 operation?

4 hours ago, faziten said:

which zotac? 

Mining wrecks VRM, and only the gpu if you apply too much voltage. Of course you can forget about mining with gpu's overclocked 24/7 without side effects on the long therm. Remember you are running out of spec. Also remember memory overclock does not affect mining whatsoever. So drop it to stock or flash a custom bios with underclocked VRAM for mining only, so you can redirect the resources to GPU only. 

 

Life span, is given by temperature, voltage (highly suggested to under volt) and PSU quality. 

 

Don't do it with a 980ti not because it might die, just because it's nonsense the money spent vs money earned. 

Mining is detrimental to VRMs health only of VRMS that are not actively cooled? Asus strix ones are I believe actively cooled

 

With regard to VRAM being not so important for mining I got my Rx580 from 24 MH up to 28 MH with just overclocking VRAM, on the other hand overclocking the core didn't improve my hashrate on Ethereum.

[Mooshi]

$6...a day...that's less than a min wage McDonald's worker makes an hour. To kill a Ti card faster than reg life expectancy? Why?

[Exaco]

Well if it would survive atleast 2 months i have nothing to lose, also it would not be 24/7 acutally, while i am gaming or something i will turn the miner off or just give few hours a day to cool it, also zec doesn't affect the vram much, the miner itself has reduced vram usage, my desktop doesn't even freeze during mining, so most likely its not using the vram much. Also if it dies i still have a warranty for around a year that should not be voided imo ( who cares if i play 24/7, render CGI 24/7 or mine 24/7 ).

It's ZOTAC AMP! Omega which is factory overclocked.
PSU is OCZ ZS 650W 80+

Also since it's loud as a vacuum cleaner i am forced to temp limit it with custom fan curve forcing the gpu to underclock itself, so the TDP doesn't even reach 75%. Also what about water cooling, would that increase the lifespan ( almost to the idle level ) if lets say the temps wont exceed 50c ?

[faziten]

14 hours ago, AgusAlexander said:

74 too high for a GPU? Those are normal temps are you saying this because of the 24/7 operation?

Mining is detrimental to VRMs health only of VRMS that are not actively cooled? Asus strix ones are I believe actively cooled

 

With regard to VRAM being not so important for mining I got my Rx580 from 24 MH up to 28 MH with just overclocking VRAM, on the other hand overclocking the core didn't improve my hashrate on Ethereum.

You are talking about different GPU's, different manufacturers, different VRM, different power draw, different brand! Nothing in common here. 

 

It may be the same algorithm but i does not mean it will behave the same on every within the same brand, let alone the other differences. 

[Simon771]

Buy new GTX 1070 for 380€ and mine with it for 3 years.

If it will die in that time, you will get new one for free, because of warranty. 

They can't prove that you were mining, and as far as I know it's not illegal to mine with GPUs.

 

Not to mention that 75°C on GPU is more than safe temp. 

 

Most likely your fans on GPU will wear off balance and will make some weird noise. But then again ... warranty!

[Delicieuxz]

Has anyone had a GPU die while mining? I mined BTC for a while years ago with 6x super hot AMD 6990 cards running 24/7. None of them died on me, and I eventually sold them off to another miner.

[AgusAlexander]

3 hours ago, faziten said:

You are talking about different GPU's, different manufacturers, different VRM, different power draw, different brand! Nothing in common here. 

 

It may be the same algorithm but i does not mean it will behave the same on every within the same brand, let alone the other differences. 

I know they differ, but most medium-high end gpus this days can go up to high 70s by just gaming.. About vram that is true: it depends, I was just pointing it out because on every forum I read what you said and in my experience it differs so I wanted to get corrected if I was wrong. 
What you said about vrm, is that for every vrm? or just for non-actively cooled ones?

[Vellinious]

3 hours ago, Delicieuxz said:

Has anyone had a GPU die while mining? I mined BTC for a while years ago with 6x super hot AMD 6990 cards running 24/7. None of them died on me, and I eventually sold them off to another miner.

Those old AMD GPUs don't mind the heat.  Maxwell doesn't perform well when it's running hot.  The cooler they are, the better they run.

 

18 hours ago, AgusAlexander said:

74 too high for a GPU? Those are normal temps are you saying this because of the 24/7 operation?

Mining is detrimental to VRMs health only of VRMS that are not actively cooled? Asus strix ones are I believe actively cooled

 

With regard to VRAM being not so important for mining I got my Rx580 from 24 MH up to 28 MH with just overclocking VRAM, on the other hand overclocking the core didn't improve my hashrate on Ethereum.

Yes, 24/7 operation at those temps will put a hurting on that GPU.  Will it die in 6 months?  I don't know the answer to that, but I sure wouldn't be surprised to see it bite it in 2.

[Exaco]

In my case the 980Ti reaches 70c whithin a minute on almost any game including CS:GO and that's with custom fan curve to reduce the heat, stock curve would reach the stock temp limit ( 83c ) in like 2-3mins on newer games. 

So i am kinda wondering why they did it such way and plus they give 5year warranty ? Maybe because its kinda heat resistant compared to other cards ? Average user would not know what is Afterburner or something, so they would just play games w/o even reading the temps and it would be running like on 83c nonstop and that should kill the card very often for many ppl making the Zotac lose the profit by sending new cards 1 or 2 times per that warranty period.

[faziten]

On 6/12/2017 at 7:14 PM, AgusAlexander said:

I know they differ, but most medium-high end gpus this days can go up to high 70s by just gaming.. About vram that is true: it depends, I was just pointing it out because on every forum I read what you said and in my experience it differs so I wanted to get corrected if I was wrong. 
What you said about vrm, is that for every vrm? or just for non-actively cooled ones?

active cooling VRM have close to zero margin of failure due to high temperatures only. It's more common to see bad soldering than actively cooled VRM failing because of overheating. 

[HardcoreBobbyD]

So after half year past what happened with your GPU, is it working yet? ETH was mooning yesterday, haha.

[Mick Naughty]

Been running 1080’s since release. No issues. Kinda the same as gaming for several years. Would easily last a year straight if you maitain tim and temp. No reason it wouldn’t last longer. 

[Metaljacket]

Unless you running 85C 24/7 without proper cooling and undervolting/underclocking It's not that much risk, heat is still the real killer, and even then the fans are still likely to fail first. I'm mining some Electroneum right now with Asus RX 470 strix OC, underclocked 1140/2000, undervolted from 120W-62W. Sitting at 65C and still getting 800h/s with custom bios strap and an extra 500h/s from ryzen 1600 @ 3.8Ghz. Even games very nice with the settings u

[BuckGup]

You are fine. My 7950 has been running at 83C 24/7 for over 3 years now and my 270X is at 2 and a half years at 77C. 

[dalekphalm]

On 6/11/2017 at 1:04 PM, Exaco said:

Hey guys, how much of lifespan decrease i can expect if i leave my Zotac 980Ti at fixed 75c temp limit for 24/7 ?
Its running at 1455Mhz Core and 3304Mhz Mem clock.

I mined on my Gigabyte HD 7950 for over year, and it ran for years after I stopped mining. I eventually sold it.

 

It will certainly add extra wear to the of the fans. But at 75C tops (I'd try to aim for a lower temp if possible), the GPU itself won't be hurting too badly.

[iiNNeX]

I need to get in on this, I have an idle 1080ti for most of the day that could be put to good use after some under-volting. It's a Strix so temps on VRMs should be good.

[dalekphalm]

5 hours ago, iiNNeX said:

I need to get in on this, I have an idle 1080ti for most of the day that could be put to good use after some under-volting. It's a Strix so temps on VRMs should be good.

Make sure to check your electricity rates, and compare that against daily potential profit from your 1080 Ti.

[iiNNeX]

19 minutes ago, dalekphalm said:

Make sure to check your electricity rates, and compare that against daily potential profit from your 1080 Ti.

Yeah will do, electricity is pretty cheap here though, especially compared to places like Germany and miners there still make a killing haha.