RMA being one of the best [strike]expert[/strike] advices, I'll get another GPU next week thanks to my contact at the store. The troubling part for me is the PCIe damage because i don't want to use my other PC during ~2-3months of motherboard RMA since everything works fine (not tested gaming yet) by reseating the gpu on another slot. Troubling because i don't know how affected the motherboard could be. And that's where i need your input guys. What caused the frying ?
Rig (assembled circa Q1 2013) :
OS: Windows 7
Case: Thermaltake level 10 GT
Motherboard: Asus Rampage IV Extreme PCIe3 (X97 LGA2011)
CPU: i7 3970x @4.5Ghz
GPU: Asus GTX Titan
PSU: Cooler Master 1300W SPH (gold)
RAM: 16GB G.skill
Avermedia PCIe x1 card
new GPU: GTX 980ti reference design
picture of the motherboard slots :
Next i'll list the slot positioning of the stages before, during and after the 980ti's frying and explain them below.
stage 1 - before :
slot 1 (x16): GTX Titan
slot 5 (x1) : Avermedia card
stage 2 - during:
slot 1 (x16): GTX 980ti
slot 3 (x8): Avermedia card
slot 4 (x16): GTX Titan (dedicated PhysX)
stage 3 - after:
slot 3 (x8): Avermedia card
slot 4 (x16): GTX Titan
Stage 1
Everything working fine, minus the occasional "driver stopped working" while playing demanding dx11 games during the last 2 nvidia updates (because strangely i can't remember crashing during a dx9 moba game i like to play and buy skins for). This setup worked for 2+ years.
Stage 2
First i just thought about replacing the GPU. But then after looking up dedicated PhysX gpu benchmarks i kept the titan inside and gave it a go. I had an sli with GTX 580 cards without issues (similar TDP values as titan and 980ti) before the titan so i had no worries about the power consumption. So first thing i did after updating the drivers and configuring my monitors back to 144hz was test witcher 3 on 1 monitor (Full HD) with everything maxed out. Graphics looked nice and i had 60-80 fps in novigrad with hairworks fully enabled ! hooray let's run arou...oops "driver stopped working" after 10minutes. Ok no problem. I know the only other thing running is msi afterburner and i read it can cause this kind of driver behaviour. Once MSI AB closed, i go back to happily running around the city for 15 minutes and screens go black and pc shuts down. The intense smell of burned components made the vision i had of me playing all those classics at 144fps nvidia surround....darken. And i thought modern gpus downclock themselves to not overheat ?
What happened :
The smell came from the gpu. I knew a smell so intense meant the card was dead, but i still had to find out what else could be affected. After trying to power the PC, the LED didn't light up and the fan didn't spin on the 980ti (the titan did), then followed a long beep and three short ones (means no vga detected on this motherboard). Next i removed the 980ti and reseated the titan on slot 1 and that didn't work either. Removing the titan and seating the 980ti on slot 4 gave the same result, dead card and beeps. Seating the titan on slot 4 booted to windows and that seemed to go fine. It's suggested to use the additional 6pin connector on the motherboard when going for quad sli, and that's something i tried to power cycle the titan on slot 1 again. Oddly enough (?) the card was detected in the bios but as x8 and could boot to windows. I didn't go further than the login since that was kind of an odd setup to be able to use the slot in my opinion. Was the slot 1 conflicting with the slot 3 Avermedia card and thus showing x8 ? (Unfortunately i never removed the card from slot 3 after stage 2, but i will test that after writing this). Were the pins damaged by the 980ti ? Anyhow, the sheer fact that nothing is detected on slot 1 without the help of the 6pin PCIe backup is a clear sign of a negative impact on the motherboard.
Stage 3
The titan is back on slot 4 where it's detected as x16 in the bios and here is the nvidia inspector :
I fear it could have affected the overall stability of my system and your insightful opinions on fried graphics cards is welcome guys. If somebody has facts that this is purely related to a faulty GPU and no other GPU can get damaged that would be a big relief. I plan on using slot 6 for the new 980ti and slot 4 for the titan and slot 3 for the avermedia card if keeping this motherboard is an option.
Additional information
- there were no differences software-wise between stage 1 and stage 2
- tasks / programs i suspect could have an impact that were running:
- - msi afterburner messing up the way the cooler works ?, but i closed it after the typical nvidia "driver stopped working" crash that happens since the witcher driver
- - geforce experience messing up the way the cooler works ?
- other tasks / programs running
- - HWinfo (in retrospect, i should have logged the data...)
- - DAEMON Tools, Skype, Canon IJ Network Scan utility, Steam, Avast, Xonar Essence One Information
- this same setup housed 2 gtx 580s in sli without psu problems in the past (during 1 intense month)
- my motherboard has 2 bios configurations accessible via a mobo button and i tested both (oc and default) on every reseat of stage 2
- i interchanged the 6pin and 6+2pin gpu connectors on every reseat
- before seating the new gpu i used a high pressure can to clean the dust inside the case, could dust particles have damaged the PCIe pins and therefore overheat the gpu ? there were no big chunks of dust though
- i still have much older gpus like a 460 or a gt9800 i could use for risky testing, but what could the tests consist of ?
....i wonder if all this makes sense and if people would read it. Thank you if you do ! Oh and i never mentioned the 980ti brand because i thought all reference cards had the same components and it didn't matter ? It's an EVGA GTX 980ti (i'll edit things if there aren't clear)
Rig (assembled circa Q1 2013) :
OS: Windows 7
Case: Thermaltake level 10 GT
Motherboard: Asus Rampage IV Extreme PCIe3 (X97 LGA2011)
CPU: i7 3970x @4.5Ghz
GPU: Asus GTX Titan
PSU: Cooler Master 1300W SPH (gold)
RAM: 16GB G.skill
Avermedia PCIe x1 card
new GPU: GTX 980ti reference design
picture of the motherboard slots :
Next i'll list the slot positioning of the stages before, during and after the 980ti's frying and explain them below.
stage 1 - before :
slot 1 (x16): GTX Titan
slot 5 (x1) : Avermedia card
stage 2 - during:
slot 1 (x16): GTX 980ti
slot 3 (x8): Avermedia card
slot 4 (x16): GTX Titan (dedicated PhysX)
stage 3 - after:
slot 3 (x8): Avermedia card
slot 4 (x16): GTX Titan
Stage 1
Everything working fine, minus the occasional "driver stopped working" while playing demanding dx11 games during the last 2 nvidia updates (because strangely i can't remember crashing during a dx9 moba game i like to play and buy skins for). This setup worked for 2+ years.
Stage 2
First i just thought about replacing the GPU. But then after looking up dedicated PhysX gpu benchmarks i kept the titan inside and gave it a go. I had an sli with GTX 580 cards without issues (similar TDP values as titan and 980ti) before the titan so i had no worries about the power consumption. So first thing i did after updating the drivers and configuring my monitors back to 144hz was test witcher 3 on 1 monitor (Full HD) with everything maxed out. Graphics looked nice and i had 60-80 fps in novigrad with hairworks fully enabled ! hooray let's run arou...oops "driver stopped working" after 10minutes. Ok no problem. I know the only other thing running is msi afterburner and i read it can cause this kind of driver behaviour. Once MSI AB closed, i go back to happily running around the city for 15 minutes and screens go black and pc shuts down. The intense smell of burned components made the vision i had of me playing all those classics at 144fps nvidia surround....darken. And i thought modern gpus downclock themselves to not overheat ?
What happened :
The smell came from the gpu. I knew a smell so intense meant the card was dead, but i still had to find out what else could be affected. After trying to power the PC, the LED didn't light up and the fan didn't spin on the 980ti (the titan did), then followed a long beep and three short ones (means no vga detected on this motherboard). Next i removed the 980ti and reseated the titan on slot 1 and that didn't work either. Removing the titan and seating the 980ti on slot 4 gave the same result, dead card and beeps. Seating the titan on slot 4 booted to windows and that seemed to go fine. It's suggested to use the additional 6pin connector on the motherboard when going for quad sli, and that's something i tried to power cycle the titan on slot 1 again. Oddly enough (?) the card was detected in the bios but as x8 and could boot to windows. I didn't go further than the login since that was kind of an odd setup to be able to use the slot in my opinion. Was the slot 1 conflicting with the slot 3 Avermedia card and thus showing x8 ? (Unfortunately i never removed the card from slot 3 after stage 2, but i will test that after writing this). Were the pins damaged by the 980ti ? Anyhow, the sheer fact that nothing is detected on slot 1 without the help of the 6pin PCIe backup is a clear sign of a negative impact on the motherboard.
Stage 3
The titan is back on slot 4 where it's detected as x16 in the bios and here is the nvidia inspector :
I fear it could have affected the overall stability of my system and your insightful opinions on fried graphics cards is welcome guys. If somebody has facts that this is purely related to a faulty GPU and no other GPU can get damaged that would be a big relief. I plan on using slot 6 for the new 980ti and slot 4 for the titan and slot 3 for the avermedia card if keeping this motherboard is an option.
Additional information
- there were no differences software-wise between stage 1 and stage 2
- tasks / programs i suspect could have an impact that were running:
- - msi afterburner messing up the way the cooler works ?, but i closed it after the typical nvidia "driver stopped working" crash that happens since the witcher driver
- - geforce experience messing up the way the cooler works ?
- other tasks / programs running
- - HWinfo (in retrospect, i should have logged the data...)
- - DAEMON Tools, Skype, Canon IJ Network Scan utility, Steam, Avast, Xonar Essence One Information
- this same setup housed 2 gtx 580s in sli without psu problems in the past (during 1 intense month)
- my motherboard has 2 bios configurations accessible via a mobo button and i tested both (oc and default) on every reseat of stage 2
- i interchanged the 6pin and 6+2pin gpu connectors on every reseat
- before seating the new gpu i used a high pressure can to clean the dust inside the case, could dust particles have damaged the PCIe pins and therefore overheat the gpu ? there were no big chunks of dust though
- i still have much older gpus like a 460 or a gt9800 i could use for risky testing, but what could the tests consist of ?
....i wonder if all this makes sense and if people would read it. Thank you if you do ! Oh and i never mentioned the 980ti brand because i thought all reference cards had the same components and it didn't matter ? It's an EVGA GTX 980ti (i'll edit things if there aren't clear)
. Good luck.
Facebook
Twitter
Instagram