SLI Crash to Black PLEASE HELP!

Hardware Hub GPU
1

I've been trying to fix this for weeks now and i'm beyond fed up. When SLI is enabled and i try to play a game my system crashes to black (no blue screen) and reboots it's self. For what ever reason it seems to happen when something tries to load. like when i'm in Far Cry 3 and try to fast travel but it also happens when I'm just playing too.

This is my system

CPU Intel 3770k
Motherboard ASUS Maximus V Extreme
RAM G.SKILL Trident X 16GB (8GBx2) 2400Mhz
GPU 2x ASUS 970 SRIX
Case Corsair 600T Arctic White
Storage 256GB Samsung 840 Pro and 2 storage HDDs
PSU 850W Seasonic 80 Plus Gold
Display(s) ASUS PB278Q 27" 1440p
Cooling Corsair h100i
Keyboard Corsair K70, Cherry MX Red
Mouse Mad Catz M.M.O. 7
Sound Objective 2
Operating System Windows 7

Nothing is overclocked.

Unless it's damaged my power supply should be more than enough to run my rig. I tested both graphics cards individually and they both work fine alone. I had the power daisy chained to them and thought that might be the problem so i ran separate power to them. I thought that fixed it because I was able to game on both cards for that whole day without a crash but now it's back to crashing.

I installed an older driver and updated my BIOS to the latest version but it didn't help.

I'm pretty sure I have a dead PCIE slot but nether of the cards are in it. could this be a PCIE buss problem?

I've also noticed my temps are a little high on my CPU running around 50C at idle and around 75C in game I think. I'm going pull my H100i to clean the radiator and replace the thermal grease but I kind of doubt that's my problem unless running both cards just puts too much heat threw the radiator. They're STRIX cards so they exhaust into the case.

I don't have another computer to test them in or another PSU to try.

I've had These cards since about the time they came out and never had a problem until recently.

2

Which PCI-E slots are you using? It's recommended to use the 8x/8x slots for SLI (1 and 3).

3

You need to test your system part by part. Put in one stick of RAM and one GPU in the top PCIE slot. If that works test the second card in the top slot. If both of those work out put them back in SLI. If it goes sour it could also be your SLI bridge. If your CPU is stock and you have an H100i and the idle temp is 50C there is an issue. If you were messing around and caused a static discharge to the board you could have fried something (i had a 3 pin fan spark while plugging it into the motherboard while it was on and my temps skyrocketed and the board went bad quick. Thats a good starting point, let us know what happens. Honestly, anytime I have a HUGE issue like this seems to be, I format and clean install everything and if the problem continues I start to troubleshoot (reinstalling Windows takes less time than troubleshooting sometimes).

4

Sounds like a motherbourd issue to me.
Things to try:

  • Swap ram into other paired slots & run @ 1600
  • run memtest86
  • turn all the C states to off in the bios
  • have a look at the pcie lane config in the bios and make sure that its all honky dory
  • contact asus directly
  • test in linux
  • check that the pcie power cable ends have good contact points inside
    • <change out the sli ribbon (unlikely but just another thing to eliminate)

Hope you have some luck mate.
5

i'm running on 1 and 3

6

i haven't tested ram but i have tested each card solo and they're both fine alone.

7

what are C states and how do i look at the PCIE lane configs?