Sign in with
Sign up | Sign in
Your question
Solved

Gigabyte GTX 780 ti GHz edition game crash

Tags:
  • Crash
  • Drivers
  • Gigabyte
  • Games
  • Graphics
  • Gtx
Last response: in Graphics & Displays
Share
May 29, 2014 4:56:37 AM

Hi everyone,

I just purchased the gigabyte 780 ti GHz edition 2 days ago, fresh installed the drivers (removing old drivers etc completely) and started to play around with the card.
Day 01:
Tomb Raider ~ No crashes
Crysis 2 ~ Crashed after 2 minutes of the game
Metro last light ~ Crashed after 2 minutes (multiple tries)
3D Mark futuremark ~ No crashes

I should have researched for this card as everyone is having trouble with Gigabyte's 780 ti and even 780s. But the problem is mine is not as severe as others.

Day 02

Tomb Raider ~ 1 hour game play - No crashes
Crysis 2 ~ No crashes
Crysis 3 ~ No crashes (yet disappointed at fps 40 avg with everything high.... I don't know if this is what a 780 ti OCd should get?
Metro last light ~ crashed after 5 minutes
Heaven 4.0 ~ 30 minutes everything extreme etc no crashes or anything
3D Mark futuremark ~ No crashes

Since everyone either RMAd their cards or refunded. Now I'm wondering if I should do the same or nothing is wrong with my card?
Only the metro LL is giving me a headache

I had GTX 680 SLi before and yet no such problem in any of those. Metro LL running smoothly with 120 fps without SSAA

System:
Intel i7 3820 @ 4.5 GHz
16 GB GSkill aries @1600 MHz
Samsung Evo 250GB SSD
Gigabyte GTX 780 ti GHz Edition
SilverStone 1000w PSU

More about : gigabyte gtx 780 ghz edition game crash

a b \ Driver
May 29, 2014 4:59:47 AM

I'd double check settings and temps and if everything is normal, then I'd rma the card.
m
0
l
May 29, 2014 5:03:21 AM

i7Baby said:
I'd double check settings and temps and if everything is normal, then I'd rma the card.


Thx for your response. Max temp recorded was in Heaven 4.0 which was 75 C stable. I don't know what you mean by the settings. Game settings i've tried everything. Didn't touched anything related to gpu volt or oc etc
m
0
l
Related resources
a b \ Driver
May 29, 2014 5:04:29 AM

Voltage, power, fan profile, etc etc
m
0
l
May 29, 2014 5:05:17 AM

taimur_111 said:
Hi everyone,

I just purchased the gigabyte 780 ti GHz edition 2 days ago, fresh installed the drivers (removing old drivers etc completely) and started to play around with the card.
Day 01:
Tomb Raider ~ No crashes
Crysis 2 ~ Crashed after 2 minutes of the game
Metro last light ~ Crashed after 2 minutes (multiple tries)
3D Mark futuremark ~ No crashes

I should have researched for this card as everyone is having trouble with Gigabyte's 780 ti and even 780s. But the problem is mine is not as severe as others.

Day 02

Tomb Raider ~ 1 hour game play - No crashes
Crysis 2 ~ No crashes
Crysis 3 ~ No crashes (yet disappointed at fps 40 avg with everything high.... I don't know if this is what a 780 ti OCd should get?
Metro last light ~ crashed after 5 minutes
Heaven 4.0 ~ 30 minutes everything extreme etc no crashes or anything
3D Mark futuremark ~ No crashes

Since everyone either RMAd their cards or refunded. Now I'm wondering if I should do the same or nothing is wrong with my card?
Only the metro LL is giving me a headache

I had GTX 680 SLi before and yet no such problem in any of those. Metro LL running smoothly with 120 fps without SSAA

System:
Intel i7 3820 @ 4.5 GHz
16 GB GSkill aries @1600 MHz
Samsung Evo 250GB SSD
Gigabyte GTX 780 ti GHz Edition
SilverStone 1000w PSU



Hi
Remove the OC of the CPU and test few days .....

I had some random problems like you have and i`ll reset mine to default speeds :) 
m
0
l
May 29, 2014 5:43:39 AM

Nikolay Savov said:
taimur_111 said:
Hi everyone,

I just purchased the gigabyte 780 ti GHz edition 2 days ago, fresh installed the drivers (removing old drivers etc completely) and started to play around with the card.
Day 01:
Tomb Raider ~ No crashes
Crysis 2 ~ Crashed after 2 minutes of the game
Metro last light ~ Crashed after 2 minutes (multiple tries)
3D Mark futuremark ~ No crashes

I should have researched for this card as everyone is having trouble with Gigabyte's 780 ti and even 780s. But the problem is mine is not as severe as others.

Day 02

Tomb Raider ~ 1 hour game play - No crashes
Crysis 2 ~ No crashes
Crysis 3 ~ No crashes (yet disappointed at fps 40 avg with everything high.... I don't know if this is what a 780 ti OCd should get?
Metro last light ~ crashed after 5 minutes
Heaven 4.0 ~ 30 minutes everything extreme etc no crashes or anything
3D Mark futuremark ~ No crashes

Since everyone either RMAd their cards or refunded. Now I'm wondering if I should do the same or nothing is wrong with my card?
Only the metro LL is giving me a headache

I had GTX 680 SLi before and yet no such problem in any of those. Metro LL running smoothly with 120 fps without SSAA

System:
Intel i7 3820 @ 4.5 GHz
16 GB GSkill aries @1600 MHz
Samsung Evo 250GB SSD
Gigabyte GTX 780 ti GHz Edition
SilverStone 1000w PSU



Hi
Remove the OC of the CPU and test few days .....

I had some random problems like you have and i`ll reset mine to default speeds :) 


Hi Nikolay, Thx for your response. That is a good idea, but I doubt that because I have been running the cpu Ocd from last 1.5 years now without any issues or temp problem, max temp gets 60 C cooled with Noctua N14 cooler. But I will try anyway
m
0
l
June 3, 2014 3:49:06 AM

Okay, so from past 4 days I've been going crazy.
1. Did a fresh install of OS 8.1 64
2. Used old Nvidia drivers and current
3. Removed any CPU OC.

btw, messing with my CPU OC made my OS corrupted! Have never seen anything like this happened before.

All my games crashes. Crysis 3 crashes with DirectX error, Metro LL crashes with Nivida driver not responding, same with crysis 2.

But the MOSSTTTT annoying part is why the FFF does the benchmarks work perfectly fine ????

My crazy local computer store people did not accepted my RMA, as they have also tested with benchmarks and asked me to get my system checked (for $50 ffs) .

Tried contacting Gigashit, they do not reply nor respond. Anyhelp is appreciated atm.

Should i mess around with the bios of the card? If so then how?
m
0
l
a b \ Driver
June 3, 2014 4:44:10 AM

Try another card. Try your card in another pc.

Check temps, voltages, clocks, fans using HWInfo

Post your whole build - mobo, case, os
m
0
l
June 3, 2014 5:53:41 AM

i7Baby said:
Try another card. Try your card in another pc.

Check temps, voltages, clocks, fans using HWInfo

Post your whole build - mobo, case, os


details

m
0
l
a b \ Driver
June 3, 2014 5:59:26 AM

Only one screen of Sensors. It shows the gpu is ok temp wise but overclocked (hitting 1228MHz).

Can you post the other 2 screens? I want to see cpu temps, voltages etc and psu voltages.
m
0
l
June 3, 2014 6:06:52 AM

i7Baby said:
Only one screen of Sensors. It shows the gpu is ok temp wise but overclocked (hitting 1228MHz).

Can you post the other 2 screens? I want to see cpu temps, voltages etc and psu voltages.


Yup, thats when i was running Heaven 4.0 it reaches 1411MHz at most.

More images
m
0
l
a b \ Driver
June 3, 2014 4:33:20 PM

They show nothing wrong.

Your cpu overclocked (4.5GHz)and warm/hot 55C. What cooler are you using?

PSU voltages OK

Case fans and cpu fans are running a bit too slow though - less than 1000rpm
Can you speed them up?
m
0
l
June 4, 2014 2:42:29 AM

i7Baby said:
They show nothing wrong.

Your cpu overclocked (4.5GHz)and warm/hot 55C. What cooler are you using?

PSU voltages OK

Case fans and cpu fans are running a bit too slow though - less than 1000rpm
Can you speed them up?


I'm using Noctua NH-D14 cooler. The fan is very annoying, I don't know how to increase the RPM. In BIOS highest is 600.

I solved the issue of the graphic card and I'm darn happy! :) 
Solution: NEVER buy gigabyte 780 ti again. I went back to my local store fully mad and bad, same tech guy met and apologised for not believing and accepting my RMA since after me couple of more people came in with similar issues. They offered me a different vendor. I went with EVGA (<3 <3)
m
0
l
June 4, 2014 2:50:39 AM

I believe I have messed up my setting a lot which is why I got some problem even with the EVGA card:

1. Crysis 3 crashed in 5 minutes, not just crashed but BSOD came with hard freeze and restarted my computer which is completely different then the crashes from gigabyte I believe i messed alot of settings which might've caused this.

2. Metro LL: I played the game for 1.5 hours without any issues (only lost 2 fps from gigabyte to Evga's card since gb is highly OCd) but I got the strangest restart. It just restarted automatically without any warning or anything.
I checked the event log and the error message was related to loss of power. I searched MS sites for solution, one guy said due to 2 audio drivers (ATI and default) was conflicted which is why he got his restarts. But I don't have a second driver as i never install the Nvidia HD audio.

But in Device manager/Sound, video and game controller its still there(nvidia audio). Disabled it but haven't tried any games yet.

this is the error message: http://imageshack.com/a/img837/3650/9vxb.png

what do you think ?
m
0
l
a b \ Driver
June 4, 2014 4:38:18 AM

OK. So you're basically saying there is no change from the Gigabyte card to the EVGA card?

The error message doesn't help.

Can you post the DUMP file contents? In c:\ Its just a text file so copy and paste. We can see if it gives any clues.
m
0
l
June 4, 2014 6:54:08 AM

i7Baby said:
OK. So you're basically saying there is no change from the Gigabyte card to the EVGA card?

The error message doesn't help.

Can you post the DUMP file contents? In c:\ Its just a text file so copy and paste. We can see if it gives any clues.


Ummmm I think it's not as bad as the gigabyte because i would get crashes on any games: dota2, cs-go, crysis2,3, metro LL and a few more in gigabyte. Metro LL lasted more than an hour on evga while on gigabyte max i could play was 10 minutes :/ 


I can't open the file, located at: c/windows/minidump
format of the file is dmp.
m
0
l
June 9, 2014 8:22:25 AM

My system is getting crazy now.

It was running fine (didn't tested much games except Metro LL for couple of days and ran smoothly) but today all of my games cannot run at all.
m
0
l
a b \ Driver
June 9, 2014 7:54:58 PM

Go back over this thread and provide any info we've asked for that you haven't yet provided eg OS, PSU
m
0
l
June 10, 2014 1:23:10 AM

i7Baby said:
Go back over this thread and provide any info we've asked for that you haven't yet provided eg OS, PSU


PSU: 1000w Zalman
OS: 8.1 64 bit

i cannot open the dump file that you asked before, .dmp format
m
0
l
a b \ Driver
June 10, 2014 1:27:50 AM

Is there one in c:\? DUMP file?

Which model Zalman? HP, XT, ....? I'm looking at manufacturer, certification and reviews
m
0
l
June 10, 2014 1:37:09 AM

i7Baby said:
Is there one in c:\? DUMP file?

Which model Zalman? HP, XT, ....? I'm looking at manufacturer, certification and reviews


The dump file is located at: C:/Windows/minidump/ sumnumbers.dmp

.PSU click here
m
0
l
a b \ Driver
June 10, 2014 2:03:50 AM

The Zalman 1000HP is made by Enhanced and certified 80plus (no rating though). It had some good reviews.

Toms - http://www.tomshardware.com/answers/id-1804779/power-su... - puts it in Tier 3 - not ideal for serious overclocking


That's about as much as I can gather from the info you've sent -

Hardware and settings seem OK

Try -

Diskcheck

System Restore to before you started getting these problems.

Memtest on your ram

Remove all overclocks, do not overclock again with this power supply.

Try the card in another pcie slot

Work out how to increase your Noctua fan speed.

Did you tell me what motherboard you got?



m
0
l
June 10, 2014 2:13:12 AM

i7Baby said:
The Zalman 1000HP is made by Enhanced and certified 80plus (no rating though). It had some good reviews.

Toms - http://www.tomshardware.com/answers/id-1804779/power-su... - puts it in Tier 3 - not ideal for serious overclocking


That's about as much as I can gather from the info you've sent -

Hardware and settings seem OK

Try -

Diskcheck

System Restore to before you started getting these problems.

Memtest on your ram

Remove all overclocks, do not overclock again with this power supply.

Try the card in another pcie slot

Work out how to increase your Noctua fan speed.

Did you tell me what motherboard you got?



mobo is asus P9 X79 pro. I'll remove the oc of cpu. Will try the memtest. I cannot system restore as it was installed a few days ago. Recently I removed Nvidia driver with help of guru3d's driver uninstaller and reinstalled. Will update if the problem comes again.
Also one more thing, My cpu (i7 3820) is not PCIE 3.0 but with a hack in regedit it gets to 3.0 (around 5-10 fps improvement). I always changed that to 3.0 but this time i didn't changed it. Will see if it has something to do with that.

m
0
l
a b \ Driver
June 10, 2014 2:29:38 AM

Why use guru3d? GeForce Experience asks if you want a clean install. Before that I just used Uninstall Programs.

Why try for pcie3.0? No cards saturate lanes at pcie2.0x8. Any of this can cause problems. I saw the same hack and ignored it.

The Asus UEFI will allow you to change cpu fan speed. See your mobo manual. Have you got the Noctua cooler fans in the cpu fan header on the mobo?

I've got x79. I believe in the KISS principle - Keep It Simple Stupid!

But you can always wipe your hard drive and start again.
m
0
l
June 10, 2014 2:43:43 AM

Well I'm desperate to try anything that would work out, reading forms and got the suggestion to use the guru3d app for uninstalling. Anyway I've used geforce experience's clean install as well.

As for pcie 3.0, you get a huge bump in the fps when you do that, at the moment, I'm getting 50fps on Watch dogs All high details. With the 3.0 enabled, I was getting 58-62 easily.

The UEFI in bios? i checked that (not in detail) and as i said the max speed was 600rpm... I'll look again.

lol @ KISS principle (but I agree) :) 
m
0
l
June 10, 2014 2:48:44 AM

Which x79 mobo are you using with which cpu ? I was thinking to go for 4930k and get the 3.0 PCIE speed, what do you reckon
m
0
l
a b \ Driver
June 10, 2014 2:53:08 AM

I'm using 3930k with ASRock Extreme 3.

If I could afford it, I'd get 4930k with ASRock Extreme 6.
m
0
l
June 10, 2014 2:55:12 AM

your gpu is already enabled with pce 3.0 speed
m
0
l
June 10, 2014 2:56:41 AM

^nevermind, i just read the 3930k and it has the revision 2.0 pcie :/  i thought it was 3.0
m
0
l
a b \ Driver
June 10, 2014 2:58:32 AM

I only really use my two gtx670's for their cuda cores for video editing hardware acceleration.
m
0
l
June 10, 2014 3:03:06 AM

try the hack and see the performance boost, I'm pretty sure the hack is not harmful, I was on it for like years without any issue. Btw, is the i7 4960X any good or how much better compared to the 4930k ?
m
0
l
June 10, 2014 3:10:22 AM

I'll go with the 4930k then, will see how much performance difference between both 3820 n this.
m
0
l
June 28, 2014 5:38:14 AM

i7Baby said:
10% better perfromance for 40% more $

http://cpuboss.com/cpus/Intel-Core-i7-4960X-vs-Intel-Co...


Hi, Sorry I've ran into a small issue and thought to ask you for help...
I bought the 4930k and it is simply awesome! I OCd to 4.6GHz (gpu is still evga 780 ti) running perfectly smooth. I also bought corsair H100i thought to try water cooling. It performs better than my Noctua NH-D14.

Anyway the cpu was bought yesterday. I played games for hours without any issues. Today I thought to stress test the beast with Prime95 and OCCT. Problem is system crashes every 5 minutes in Prime95 Test and OCCT fails within 2 minutes with message: Core#0 Max value reached = 185 degrees (which is ridiculous and isn't even possible?) maximum I've seen my cpu temp in MSI AB, ASUS AI suit, RealTemp is 74 C. CPU idles at 26-30 C which is perfect. I'll post some pics.

Thanks
m
0
l
a b \ Driver
June 28, 2014 5:47:15 AM

That's what Prime 95 is for - to test the OC. If it's crashing, you have to back off on the OC. Drop Vcore .05V
m
0
l
June 28, 2014 5:47:18 AM

OCCT & HwInfo

OCCT now cannot run at all! just for 2 second and gives the error message of max value reached :( 
m
0
l
June 28, 2014 5:51:27 AM

i7Baby said:
That's what Prime 95 is for - to test the OC. If it's crashing, you have to back off on the OC. Drop Vcore .05V


I've never touched voltages, only increased the BLCK and the ratio thing to 37 and BLCK to 125 i think..
m
0
l
a b \ Driver
June 28, 2014 5:53:47 AM

Your Vcore is OK. Your core temps are too high. Must be the h100 not working properly. Check fan speeds and pump speed. If OK, re-seat the block.

Why does OCCT say your Vcore is 1.45 but HWInfo says 1.2V? Check with your UEFI. Keep Vcore less than 1.25V
m
0
l
June 28, 2014 6:04:01 AM

i7Baby said:
Your Vcore is OK. Your core temps are too high. Must be the h100 not working properly. Check fan speeds and pump speed. If OK, re-seat the block.

Why does OCCT say your Vcore is 1.45 but HWInfo says 1.2V? Check with your UEFI. Keep Vcore less than 1.25V


Thats bothering me, I mean temps never exceeds (when i monitor) 73-75 degrees. H100i seems fine i don't see any seating issue or anything, also the fan speed runs ridiculously loud and rpm gets to 2300 when under load.

I'll report back with UEFI Vcore if its more than 1.25V but just out of curosity (sorry I dont have much exp in here) is 1.25V too less for OC ?
m
0
l
June 28, 2014 10:58:27 PM

I'm not using 3820 anymore :)  got the 4930k. I really don't know how to work on with the Vcore voltage. In bios, there is CPU VCORE Voltage with 2 options, Offset Mode(Current) and manual Mode. At the moment the cpu is overclocked (4930k not 3820) at 4.445GHz and VCORE voltage is 1.404 by default like i didn't do anything. How should i change this to 1.25V ?
m
0
l

Best solution

a b \ Driver
June 28, 2014 11:43:29 PM

If its overheating then don't overclock it!

When you want to overclock look up overclocking guides for doing it first.

The H100i should keep your cpu to under 55C even under load. On idle it should be under 40C. If its not, then it hasn't been set up right.

Double check your cpu temp. Use your mobo UEFI.
Share
June 29, 2014 1:36:14 AM

I read some forms and many people's suggestion was to increase the multiplier rather than the BCLK (which was at 125 all the time). I changed that to 100 and the multiplier to 43. Seems pretty stable, Idling at 24-28 C and medium load is 45-55 C. With full load using prime 95 it gets maximum 66 C which is much better. Only the OCCT is failing. I guess there is something wrong with OCCT as the failed temps that it reports is ridiculous (150 C) lol.

Thanks for your help.
m
0
l
a b \ Driver
June 29, 2014 2:10:55 AM

No worries.
m
0
l
!