1070 Overheat Crash x(

[Cpt.Fox]

Hi guys,
my pc just crashed and i had to remove my GPU, i touched it and it was very hot, I never had this issue before, i dont know why it happened, i was gaming but nothing out of the ordinary. I tried running the PC again with the card installed but nothing, i will let it cool for a while and try again tomorrow. Why do you think this happened? I have the latest 384.76 nvdia driver but not the EVGA presicion software or anything like that . i pray to the gods that the card isnt dead or something :S.
Id appreciate your help

 

[Toothless]

Bad paste, borked fans, anything really. Sounds dead tbh.


EDIT: EVGA cards had vrm issues. RMA if you can.

 

[Cpt.Fox]

goddamnit! if its really dead and RMA was the only option, it would take quite some time since my dad bought it when he travelled to the USA and he brought it back to our homeland about6 months ago. How long would it take to ship it there and back to my country, i wonder if there is a way to RMA in Chile but i dont think so. How long does EVGA usually take to ship over new cards? I cant believe this &#$@ . Lets hope it works if i try again tomorrow

 

[Toothless]

Just contact them. We can't answer these things for you.

 

[Cpt.Fox]

guess ill have to go for option number 3 https://www.evga.com/howtorma/
however i will keep trying tomorrow and let you know if i have any success reviving it, wish me luck

 

[EarthDog]

Cards are "hot" when you touch them anyway... people they run at 70c+....

Without other information or testing, who knows if tbe card is dead. Could be anything. Did you try using the igpu and see if it turns on? Try the 1070 in a different pc? Why did you "have to remove your gpu??"

More clear info please... this is like half a thought...lol

 

[Vayra86]

I see EVGA FTW 1070 - specifically the EVGA cards with VRM issues. If you push this card the heat from the VRM section will radiate over the memory chips around the GPU die and likely go over 95C for which the VRAM is rated = boom dead card. The backplate on this card works as an insulator, not a heat conductor.

This is why I returned this card immediately and never look at EVGA again. Most notably because they let you install the thermal pads yourself and because they downplayed this issue stating the 1070 should never have the problem. Meanwhile, people could produce artifacting cards left and right.

When there's smoke...

 

[EarthDog]

Oh, the furmark thing killing the card... when normal use didnt.. gotcha.

He also didnt mention overclocking so we have to assume its stock until he tells us he 'pushed it'.

I mean, it could be dead, but, more testing is needed to confirm. Waiting on the op at this point...

 

[Vayra86]

Oh, the furmark thing killing the card... when normal use didnt.. gotcha.

He also didnt mention overclocking so we have to assume its stock until he tells us he 'pushed it'.

I mean, it could be dead, but, more testing is needed to confirm. Waiting on the op at this point...

It's summer time. GPUs run hot.

I know I'm jumping to conclusions, but I am one of those people that owned this card and saw it behave the way it did, making all my alarm bells go off. Could cook an egg on that backplate.

 

[EarthDog]

You assume ambient changes significantl/ he doesnt have AC to keep temps in the ballpark...

It was warm/hot, but within spec. Typically, a thermal shutdown prevents damage, its not like they all croak after passing a certain temp... thats why have thermal shutdown to pregent damage. He didnt see artificating, etc, just shutdown. Psu could be the issue, temps... etc.

Will wait for op...

 

[GreiverBlade]

reading EVGA does the trick ... likely not furmark ... EVGA does not need furmark to burn, specially with the 10XX series : a true "DNB" brand for me (no wonder why they give a "lifetime" warranty ... only one card i had from them didn't need it )

the current 10XX result is like that for me:
had issues and corrected them (mostly) : Asus, Gigabyte

have issue never corrected them/told the buyer to fix the problem by themselves with provided thermalpad (what? just thermalpad ... no additional heatsink handed with? ... humpf ): EVGA

never had any issues from the start (ok Micron vRAM bug ... thought rare for them, unlike others): MSI, even the cheapest 1070 from MSI have a heatsink covering vRAM/VRM
the ACX 3.0 feel cheap compared tho the Armor

from user feedback, you want a card that has no issues (reviews tend to have sample with no issues, for various reasons i guess ) you go MSI, i didn't see much complaint about them and i run one without any issues since out of the box (thought i use 376.53 drivers)


oh well, i hope the RMA process will go smoothly for the OP

i pray to the gods

[joke]will not work .... it never work ... [/joke]

Typically, a thermal shutdown prevents damage, its not like they all croak after passing a certain temp... thats why have thermal shutdown to pregent damage. He didnt see artificating, etc, just shutdown. Psu could be the issue, temps... etc.

i hope he didn't see artifacting ... after all it's EVGA we are talking about...

i doubt it's the PSU tho 550 might be a little short (not insufficient but well non OC GPU/CPU on a 550w should be fine ) i would have taken a 650w minima (hence why i run a 750w in my rig)

 

[EarthDog]

You post is weird..quoting or something. Cant make sense of the first part.

Edit: ok, i got it after a couple rereads... last i recall, furmark was the issue. Otherwise, the vrms ran in spec, but people 'felt' they were running too warm. Evga is known to have stellar customer service. It it really was as big an issue as the lemmings made it out to be, they would have done more than offer a diy thermal pads..dont you think?
https://linustechtips.com/main/topic/678970-evga-responds-to-hot-vrm-area-on-gtx-10-series/
Edit: they ran in spec with furmark too...

Anyhoo, that g2 550 is PLENTY to run both cpu and gpu overclocked to their ambient limits. Youd be lucky to hit over 400w.

 

[GreiverBlade]

You post is weird..quoting or something. Cant make sense of the first part.

Anyhoo, that g2 550 is PLENTY to run both cpu and gpu overclocked to their ambient limits. Youd be lucky to hit over 400w.

mmhhhh? the 1st part is not linked to any quote ... and just sink down EVGA for their 10XX line ...

and yep a 550w is enough i got a 750w in case i would go SLI but likely not (since on load i am not so far from the max )
OP guestimation
Load Wattage:337 W
Recommended PSU Wattage:387 W (recommended PSU is still a 650w tho, and a SuperNova G2 to add ... at last they have good PSU ... if not GPU )
my own RIG...
Load Wattage:686 W
Recommended PSU Wattage:736 W

("DNB"= Do not buy just in case ... but i am sure you did know what it mean )

 

[EarthDog]

Read my updated post plz. Does anyone refresh or see the notifications the forum provides? Lol!

IMO evga shouldnt be on anyone's DNB list. The vrm issue was blown WAY out of proportion amd their customer service is generally held in the HIGHEST regard.

So now the psu is fine, was a little short earlier? And then these links and wattages... lol. Yikes.. who cares what calcs say...what ARE you trying to say???

Those calcs overestimate a ton. People are still suckered in to buying way more than they need for psus.

 

[Cpt.Fox]

Cards are "hot" when you touch them anyway... people they run at 70c+....

Without other information or testing, who knows if tbe card is dead. Could be anything. Did you try using the igpu and see if it turns on? Try the 1070 in a different pc? Why did you "have to remove your gpu??"

More clear info please... this is like half a thought...lol

Ok guys im back, last night when it crashed ( didnt see artifacts) i tried to boot the PC with card still on but nada, so i removed the GPU and then i could boot the OS using the IGPU with no problems, so up to now everything is good except the card. I didnt try booting the PC again with the card on back then because the card was still hot, but now i could give it another shot, if you recommend doing so... I cant test the 1070 on another PC, i dont know anyone that has a good PC able to support this GPU,i would have to go to a tech store and have them test it there , but that can take days and i dont trust them people to fix it, id rather do it with your help and try to fix it with what we have.

I see EVGA FTW 1070 - specifically the EVGA cards with VRM issues. If you push this card the heat from the VRM section will radiate over the memory chips around the GPU die and likely go over 95C for which the VRAM is rated = boom dead card. The backplate on this card works as an insulator, not a heat conductor.

This is why I returned this card immediately and never look at EVGA again. Most notably because they let you install the thermal pads yourself and because they downplayed this issue stating the 1070 should never have the problem. Meanwhile, people could produce artifacting cards left and right.

When there's smoke...

I didnt know that issue with furmark, never used that before. another important thing to note ist that am not OCing the GPU nor the CPU, everything is at stock speed, and here its not summer, its winter so its even stranger that this overheat happened.
I think the EVGA 550 gold PSU is more than enough, i did the math back then when my dad was buying the parts and i was told it was plenty of wattage for single GPU build, i would have gone for 650 but my dad had little time and it was the best we could find. i set the PSU to Eco mode to save some power, maybe thats something i shouldnt have done? so far it has been running smooth and cool.

Does EVGA still give away those thermal pads? according to this forum topic https://linustechtips.com/main/topic/678970-evga-responds-to-hot-vrm-area-on-gtx-10-series/ i should have to update the bios or do an RMA an have an update replacement, is this necessary?
To tell you the truth the only reason i went for EVGA was because of the Life time warranty, i dont live in US but i have an uncle there that could help me if have issues with any component. Before this build i used an MSI GTX8800 that died 2 years ago, it lasted almost 10 years. I hope we wont have to RMA , the Import fees are not gonna be cheap and we know it would suck having to wait for this.
Pls tell me what is the next step now, im waiting for your command
thanks for you help amigos

 

[EarthDog]

Yes try it again... lol... we were waiting for that!

You have steps listed....try to find a another pc to try...try to reset the bios and see if that for some reason would help. But yes, try putting it back in. If not, id say rma time.

 

[Cpt.Fox]

Bad news x(, i installed the gpu and the pc didnt boot, so i took it out and guess what, it doesk work either . Im worried that the PSU is damaged, how can i reset the bios? I removed the battery for a second but im not sure its the only way. Pls advise on how to proceed now. im writing you from my phone.

 

[EarthDog]

Isnt there a button or jumpers in the board? Try those...

(Can see in manual how to)

 

[Cpt.Fox]

i cant seem to be able to reset the bios, I removed the battery for 10 seconds and then l alo tried the CLR CMOS bridge with a screwdriver ( coudnt find the jumper). How else can i make the PC boot again? do you think its the motherboards fault or the PSU? When i plug the Power Cord to the PSU i can hear some electric noise, so im pretty sure its the Mobo. i will try to test the GPU on a friends PC, i might be able to do that in the meantime.

 

[Vayra86]

To me it seems unlikely the PSU is damaged out of the blue now. Its possible, but not likely, also is a quality PSU. ECO mode doesn't hurt its performance, for one.

Its definitely worth testing the GPU in another system. If it works, I would continue the search on the motherboard side, you can also visually inspect it (the traces on the board, and capacitors on it, see if any of them are bulged / show corrosion).

Did anything smell funny throughout this process?

 

[Cpt.Fox]

no funny smells at all, i will try to bridge the PSU or test it with a multimeter to see if its really ok

 

[jboydgolfer]

Cards are "hot" when you touch them anyway... people they run at 70c

Yup,

this shouldn't be a surprise to anybody who owns a computer. that's why there's a sticker on many video cards and other PC components that says "warning components may be hot"

 

[Cpt.Fox]

hey guys good news! the PC is back up, but the GPU hasnt revived yet, do you think if i RMA they will send a newer version of the 1070 with thermal pads on the backplate? i hope they accept my request, they ask you to send details about the issue , what would you specify? i think i can only say that we tried every possible way to fix the card
"EVGA Guest RMA

Please enter in the issue you are having and any troubleshooting you have done. The more detail you provide at this step, the faster you will receive your RMA approval."

 

[Vayra86]

Try this

Jokes aside, specify in the e-mail what you tried and how the card is responding at this time. EVGA will do their own check anyway.

Pro tip: don't start about the motherboard and PSU. Specify system specs and stick to what the GPU did and what you did with the GPU.

 

[Cpt.Fox]

haha thanks man, they say they will respond to the RMA request in 1 or 2 bness days, wish i had found the receipt earlier. Lets hope they apply the magic of that card.
In the meantime im using the CPU IGPU, gotta say its not that bad as i suspected, you can still game a bit and watch videos