2 GTX 980s, 3 Monitors at 1440p, 60hz, Random Disconnecting from Monitors

[jfrontier]

Hello,

I have 3 Dell Gaming monitors(1440p 144hz) and wanted to make sure I can game on at least one monitor at 1440p 144hz while the other 2 are at 1440p at 60hz with no problem.

I bought a secondary 980(both of them are ASUS Strix GTX 980s) with the same clock speed and decided to use one for my main 144hz monitor and the other for my 60hz.

Though for testing purposes I put them all at 60hz for the time being(it's still crashing).

So for whatever reason, I cannot replicate it in a deterministic manner is that they will randomly disconnect, not crash. The drivers will still be fine and all apps are running but the graphics card loses the connection.

Here are my specs from Speccy:

Operating System
Windows 10 Pro 64-bit
CPU
Intel Core i5 6600K @ 3.50GHz 29 °C
Skylake 14nm Technology
RAM
32.0GB Dual-Channel Unknown (15-15-15-36)
Motherboard
ASUSTeK COMPUTER INC. Z170-A (LGA1151) 31 °C
Graphics
Dell S2716DG (2560x1440@59Hz)
4095MB NVIDIA GeForce GTX 980 (ASUStek Computer Inc) 53 °C
4095MB NVIDIA GeForce GTX 980 (ASUStek Computer Inc) 53 °C
ForceWare version: 391.35
SLI Disabled
Storage
931GB SAMSUNG HE103UJ (SATA) 26 °C
931GB Samsung SSD 850 EVO 1TB (SSD) 30 °C
2794GB TOSHIBA DT01ACA300 (SATA) 32 °C
2794GB Western Digital WDC WD30EFRX-68EUZN0 (SATA) 29 °C
Optical Drives
LITE-ON DVDRW LH-20A1L
Audio
NVIDIA Virtual Audio Device (Wave Extensible) (WDM)

 

[Sir_Lagsalot]

Do all of the monitors disconnect or just one? Try connecting the monitors to different outputs on the GPUs.

 

[jfrontier]

Do all of the monitors disconnect or just one? Try connecting the monitors to different outputs on the GPUs.

After further testing, it might be narrowed down to the PSU or GPU.

I ran a stress test on a single monitor being driven by 1 GPU(the one that I bought used that just arrived yesterday). And it crashed immediately. The monitor disconnected and I can only assume the drivers failed. I was still able to input but I am unaware if the heat temps got too hot.

I was doing this on a 750w power supply, and running this one card, not both, when this failed as well

But further reading suggests I could of fried the PSU by trying to run both cards and it not being enough.

After the stress test and crash I got artifacts on the BIOS and it freezes with a white line on startup, hinting that the GPU is faulty.

I'm not sure if I fried my card or not.

 

[Sir_Lagsalot]

Artifacts and freezing in the bios most likely means a faulty card. Do you have another system that you could test the gpu with, just to isolate it and confirm that it is what is causing the issue? Also try booting up your system with your 6600k's onboard graphics, to test if it is the gpu or psu causing the issue. If the artifacts remain, it is likely the psu. If they are gone, then the gpu.

 

[jfrontier]

Artifacts and freezing in the bios most likely means a faulty card. Do you have another system that you could test the gpu with, just to isolate it and confirm that it is what is causing the issue? Also try booting up your system with your 6600k's onboard graphics, to test if it is the gpu or psu causing the issue. If the artifacts remain, it is likely the psu. If they are gone, then the gpu.

I have booted with Onboard graphics, it boots fine, no artifacts at all. I have an identical 980 but I don't want to risk putting this in as it might also fry it as well? I'm not sure how an underpowered PSU can brick a GPU but i've read that it's possible.

Is the GPU thats artifacting/crashing too broken to fix? Would a new PSU help it?

 

[Sir_Lagsalot]

What psu to you have? Brand and model if you can.

 

[jfrontier]

What psu to you have? Brand and model if you can.

Corsair CXM CX750M

 

[Sir_Lagsalot]

What psu to you have? Brand and model if you can.

Corsair CXM CX750M

Hmmm... i've read that that psu doesn't play too well with flagships in SLI, but have never heard of it frying them. A new psu would probably be the best choice here. If the disconnections and artifacts remain an issue, then the gpu is likely at fault.

 

[Sir_Lagsalot]

What psu to you have? Brand and model if you can.

Corsair CXM CX750M

Hmmm... i've read that that psu doesn't play too well with flagships in SLI, but have never heard of it frying them. A new psu would probably be the best choice here. If the disconnections and artifacts remain an issue, then the gpu is likely at fault.

 

[jfrontier]

What psu to you have? Brand and model if you can.

Corsair CXM CX750M

Hmmm... i've read that that psu doesn't play too well with flagships in SLI, but have never heard of it frying them. A new psu would probably be the best choice here. If the disconnections and artifacts remain an issue, then the gpu is likely at fault.

I have this coming into tomorrow

https://www.amazon.com/Rosewill-Certified-Control-Semi-modular-RBR-1000MS/dp/B003J89V0A/ref=sr_1_14?ie=UTF8&qid=1522812847&sr=8-14&keywords=1000w+power+supply

I'll update if it is indeed the PSU.

 

[Sir_Lagsalot]

What psu to you have? Brand and model if you can.

Corsair CXM CX750M

Hmmm... i've read that that psu doesn't play too well with flagships in SLI, but have never heard of it frying them. A new psu would probably be the best choice here. If the disconnections and artifacts remain an issue, then the gpu is likely at fault.

I have this coming into tomorrow

https://www.amazon.com/Rosewill-Certified-Control-Semi-modular-RBR-1000MS/dp/B003J89V0A/ref=sr_1_14?ie=UTF8&qid=1522812847&sr=8-14&keywords=1000w+power+supply

I'll update if it is indeed the PSU.

Alright, good luck to you.

 

[jfrontier]

So I tested with the new power supply and it turns out it's what I think is the worst case scenario.

I put it in and tested the GPU with a relatively simple benchmark and it crashed.

Though to be sure, I put in my previous identical card(the one i've been using for 4 years) and it resulted in the error.

The PSU must have damaged both cards when they were under load. Which is strange because it wouldn't seem like running a game in 1 window at 60hz 1440p would cause massive power draw, but here we are.

 

[Sir_Lagsalot]

So I tested with the new power supply and it turns out it's what I think is the worst case scenario.

I put it in and tested the GPU with a relatively simple benchmark and it crashed.

Though to be sure, I put in my previous identical card(the one i've been using for 4 years) and it resulted in the error.

The PSU must have damaged both cards when they were under load. Which is strange because it wouldn't seem like running a game in 1 window at 60hz 1440p would cause massive power draw, but here we are.

Damn. At least you found what was causing the issue. It's too bad that it was that serious.