Is my GTX 770 dead?

[RoughJustice]

Title says the main question..... I’ve had this card for maybe 5 years now and has never given me any problems. Today I was watching twitch on one monitor and playing wow on another. All of a sudden the screen froze and glitches out in weird pixles and then my computer restarted itself and I got a blue screen with the error 0x00000116.

This is basically everything I have done since then:
First thing was I restored windows from a previous date (yesterday), once that went to restart again it gave me the blue screen once again. So I tried restoring from another restore point but now windows is telling me that there are no restore points available? I’ve never had that issue before.

Next I restart once again and go into safe mode and uninstall the gpu driver, restart and windows boots up fine so I go to Nvidia’s site to get the latest driver (mine wasn’t current). I instal that and restart once again and receive another blue screen.

By now I’m losing hope, open my tower up and pull the gpu out and clean it and repeat the last step but get another blue screen.

I took the gpu from my wife’s computer and put that in and windows loaded fine, downloaded the drivers, restarted and windows loaded back to normal.

So what else could I try to get my 770 back to life?

Computer specs:
i5 2500k
EVGA gtx 770
Asrock z77 extreme4

[ATZenith]

This happened to be, or something similar. My harddrive had fried from my GPU having hot temperatures. Was your GPU running hot? You sure it's your GPU and not sure harddrive?

[the boar]

don't do it just because i've said it, see what others have to say and read other peoples experiences first.... but the very final thing you can try for a dead graphics card.. bake it in the oven.

i kept an old 8800 GT going for an extra 6 months by giving it a cook! but its definitely a last ditch option, and you should be at the point you're ready to call it a day and buy a new one.. for obvious reasons that come along with putting your graphics card in the oven.

that said, it wont keep it alive forever if it does work. and at that point you should probably start saving up for a new one, because if the baking does revive your card, it wont be too long before it dies again.

[Will]

The life span of the card has many variables affecting it; whether it was OCd, how dusty your case interior is, how much time the PC was on, how much time it was under load, even the quality of your PSU, these things all have a part to play.

Assuming the card does still work (up until the point it fails again, naturally) I'd monitor the gpu core temperatures using a program like MSI Afterburner. Check temps. Under max load (stress test) you should see at the very most temps of around 80c if you're factory overclocked. If temps are above that, there's your problem, but fear not, it may be that applying new thermal compound to the gpu heat sink solves temperature problems. Search youtube for a guide how to do that if unsure. Try that before buying a new card. It's worth noting that a card can handle temps of even up to 90c, but if the card were to regularly sit at such temps it would kill the card way quicker. If temps are that high it could be why it failed. Could also be that it was simply the card's time to fail

If temps are not a problem then check the power connectors to the card are in securely, remove dust from the card (if any) - never hurts to take it out of the case and blow into and over it with an air duster every 6 months to a year. Still having problems? Then might be time to get a new card.

- - - Updated - - -

This happened to be, or something similar. My harddrive had fried from my GPU having hot temperatures. Was your GPU running hot? You sure it's your GPU and not sure harddrive?

How does that even happen? The core temps of the GPU are way hotter than the rest of the card. There's no way a hot GPU can somehow fry an HDD which is located elsewhere in the case. Your gpu would literally melt before any heat from it even came close to superheating the air in the case to a temperature capable of damaging an HDD, unless the HDD was quite literally touching the GPU, which would be dumb.

- - - Updated - - -

don't do it just because i've said it, see what others have to say and read other peoples experiences first.... but the very final thing you can try for a dead graphics card.. bake it in the oven.

i kept an old 8800 GT going for an extra 6 months by giving it a cook! but its definitely a last ditch option, and you should be at the point you're ready to call it a day and buy a new one.. for obvious reasons that come along with putting your graphics card in the oven.

that said, it wont keep it alive forever if it does work. and at that point you should probably start saving up for a new one, because if the baking does revive your card, it wont be too long before it dies again.

Before baking it at the very least I would try a. De-dusting it, b. refreshing new thermal paste on the heatsink, c. monitoring temps using software on the desktop under various conditions and d. check it isn't running at some dumb overclock frequency (which may have been triggered accidentally or while drunk or whatever)

[Rukario]

Change your RAM first, remove the sticks...try one at a time and see what happens.

Odds of you having RAM issues are far greater than your GPU being dead, unless you messed with it somehow(overclocking).

[Evildeffy]

I'll be caught dead before linking this again but he did something I've done multiple times (only he was a couple years late).



You can try this if you can find the exact model and BIOS of your GPU here:
https://www.techpowerup.com/vgabios/...emSize=&since=

Your own BIOS version can be found when basic MS drivers are installed with GPU-Z telling you it.

You better damned appreciate this because I'd prefer not to link JayzTwoCents ever again... the guy can do water cooling builds but he's not a damned Techie.

[potis]

I'll be caught dead before linking this again but he did something I've done multiple times (only he was a couple years late).



You can try this if you can find the exact model and BIOS of your GPU here:
https://www.techpowerup.com/vgabios/...emSize=&since=

Your own BIOS version can be found when basic MS drivers are installed with GPU-Z telling you it.

You better damned appreciate this because I'd prefer not to link JayzTwoCents ever again... the guy can do water cooling builds but he's not a damned Techie.

Realistically speaking, isnt that just because the (Custom) bios probably has different power settings 99% of the time, mostly unlocked ones and forced higher TDP?

Which simply makes the card work for awhile before whatever is cluttering and causing artifacts goes back to doing it as it breaks down?

[Evildeffy]

Realistically speaking, isnt that just because the (Custom) bios probably has different power settings 99% of the time, mostly unlocked ones and forced higher TDP?

Which simply makes the card work for awhile before whatever is cluttering and causing artifacts goes back to doing it as it breaks down?

Nope, the vBIOS versions on the TechPowerUp site in general are originals unless specifically stated.

F.ex. .. MSI has/had 3 modes on their cards that you had to use in order to reach max rated speeds that they were advertising with on the box.
My MSI R9 390X Gaming 8G f.ex. has "Silent Mode" running @ 1020MHz Core/1500MHz Memory -> "Gaming Mode" running @ 1050MHz Core/1500MHz Memory -> "OC Mode" running @ 1100MHz Core/1525MHz Memory even though the box's specs state 1100MHz/1525MHz as a default.

Yet this wasn't achieved on it's own unless you specifically OCed them or used the specific MSI App ... which later they got a lot of flak for.
So they released BIOS versions that would allow you to use the advertised speeds without the need for the App.

It could be considered a "Custom vBIOS" ... but in this matter it's official.

Same goes for the rest of the vBIOSes ... as well as possible revisions for refinements.

Though I've never seen an AMD card actually corrupts it's vBIOS I have seen it happen on nVidia cards.
I wouldn't say it happens regularly but often enough for me to recognize issues and attempt to fix it that way by re-flashing the original BIOS to attempt to fix it.

If you remember properly GigaByte's 8-series Intel boards had a lot of BIOS corruption issues over time as well which is why they implemented a non-reachable M_BIOS and B_BIOS chips on their boards to auto-reset if M_BIOS is having issues.

This issue is relatively more common in general just less on the GFX side of things.

[potis]

Nope, the vBIOS versions on the TechPowerUp site in general are originals unless specifically stated.

F.ex. .. MSI has/had 3 modes on their cards that you had to use in order to reach max rated speeds that they were advertising with on the box.
My MSI R9 390X Gaming 8G f.ex. has "Silent Mode" running @ 1020MHz Core/1500MHz Memory -> "Gaming Mode" running @ 1050MHz Core/1500MHz Memory -> "OC Mode" running @ 1100MHz Core/1525MHz Memory even though the box's specs state 1100MHz/1525MHz as a default.

Yet this wasn't achieved on it's own unless you specifically OCed them or used the specific MSI App ... which later they got a lot of flak for.
So they released BIOS versions that would allow you to use the advertised speeds without the need for the App.

It could be considered a "Custom vBIOS" ... but in this matter it's official.

Same goes for the rest of the vBIOSes ... as well as possible revisions for refinements.

Though I've never seen an AMD card actually corrupts it's vBIOS I have seen it happen on nVidia cards.
I wouldn't say it happens regularly but often enough for me to recognize issues and attempt to fix it that way by re-flashing the original BIOS to attempt to fix it.

If you remember properly GigaByte's 8-series Intel boards had a lot of BIOS corruption issues over time as well which is why they implemented a non-reachable M_BIOS and B_BIOS chips on their boards to auto-reset if M_BIOS is having issues.

This issue is relatively more common in general just less on the GFX side of things.

I didnt know this started from 700 series, thought it was more a "WTF is this shit" after 900 series with the MSI App.

The only BIOS corruption i have encountered has a MSI P67 tbh, even so it did work perfectly fine for months if you did manage to boot just once!

Which reminded me, i havent actually checked the MSI APP in awhile

[callipygoustp]

Sounds like you're video card is on it way out, seen something similar with a bad memory stick as well. You should take a look at the *.dmp files with something like WhoCrashed or, preferrably, WinDbg. Might get some information that will help you hone in on the exact issue.

[SinR]

that video was actually kinda cool. something I'll have to remember if my 980 ever dies.

Dude seems like a tool, though

[ATZenith]

The life span of the card has many variables affecting it; whether it was OCd, how dusty your case interior is, how much time the PC was on, how much time it was under load, even the quality of your PSU, these things all have a part to play.

Assuming the card does still work (up until the point it fails again, naturally) I'd monitor the gpu core temperatures using a program like MSI Afterburner. Check temps. Under max load (stress test) you should see at the very most temps of around 80c if you're factory overclocked. If temps are above that, there's your problem, but fear not, it may be that applying new thermal compound to the gpu heat sink solves temperature problems. Search youtube for a guide how to do that if unsure. Try that before buying a new card. It's worth noting that a card can handle temps of even up to 90c, but if the card were to regularly sit at such temps it would kill the card way quicker. If temps are that high it could be why it failed. Could also be that it was simply the card's time to fail

If temps are not a problem then check the power connectors to the card are in securely, remove dust from the card (if any) - never hurts to take it out of the case and blow into and over it with an air duster every 6 months to a year. Still having problems? Then might be time to get a new card.

- - - Updated - - -



How does that even happen? The core temps of the GPU are way hotter than the rest of the card. There's no way a hot GPU can somehow fry an HDD which is located elsewhere in the case. Your gpu would literally melt before any heat from it even came close to superheating the air in the case to a temperature capable of damaging an HDD, unless the HDD was quite literally touching the GPU, which would be dumb.

- - - Updated - - -



Before baking it at the very least I would try a. De-dusting it, b. refreshing new thermal paste on the heatsink, c. monitoring temps using software on the desktop under various conditions and d. check it isn't running at some dumb overclock frequency (which may have been triggered accidentally or while drunk or whatever)

I don't know how it happened tbh. My temps were getting up to 98 though. I didn't really know anything about that since the PC was given to me. But the harddrive died and I made sure to get my temps to normal levels. My computer would reset when playing Overwatch sometimes. I knew something was wrong but I didn't know what was wrong. One day it shutdown and it was the HD wasfried. I made damn sure to never be stupid again when it comes to that.