PC shuts off under GPU load + POST code 97

[namireads]

Hello reader, I am looking for help after a years worth of frustrating GPU related issues;

 

Since building my PC I randomly fail to boot and receive POST code 97. This has happened twice in the last year leading to two RMAs and one replacement card. Everything seems to work fine for a few weeks until the dreaded POST 97 returns. Today, after enjoying a perfectly working PC for ~6 months, I have noticed my screen flicker and PC shut off whilst gaming. I tested this with heaven benchmark and the same thing happens. I get about 3 tries with this until I'm hit with the POST 97 and can't boot unless I remove the GPU and use the integrated graphics.

 

Testing

1) I have tried a spare 1060 in my build which works fine in both PCIe slots.

2) I tried with/without PCIe extension cables but issue persists (I usually use extension cables for aesthetic).

3) I have tried the GPU in a different PC and that works fine (bearing in mind currently 2 RMAs deep and replacement card).

4) Please see attached image to see HW monitor  immediately before the screen turns off and POST 97 occurs, whilst Heaven Benchmark is running.

Spoiler

5) PC runs fine with no GPU installed.

 

Notes

1) Nothing is overclocked

2) Running XMP II profile

3) Win 10 64 Bit OS

4) BIOS has been updated to latest version (1402)

5) GPU has been updated with latest drivers

6) Windows updated to latest drivers

 

Specs

https://uk.pcpartpicker.com/list/p9TPKp

 

I would appreciate any help with this matter - apologies if I missed any details.

 

Many thanks

[Hairless Monkey Boy]

First thing to try is getting rid of the extension cables.

[namireads]

9 minutes ago, Hairless Monkey Boy said:

First thing to try is getting rid of the extension cables.

Many thanks for your response. I have already tried running without the extension cables but problem persists.

[Hairless Monkey Boy]

18 hours ago, namireads said:

Many thanks for your response. I have already tried running without the extension cables but problem persists.

Might have to RMA the PSU then.

[namireads]

On 6/29/2022 at 1:42 PM, Hairless Monkey Boy said:

Might have to RMA the PSU then.

Thanks. I've started the RMA process.

 

On a side note, when inspecting the PCIe power cables powering the GPU, I noticed that one of the pins had come out. I immediately thought this was the issue, pushed the pin back in and retested only to still have the problem! I also have switched to running to seperate PCIe cables directly from PSU to 2x8 pin GPU as I read this may be better for the higher powered cards (I was running a single pig-tailed PCIe before).

 

I can currently turn PC and and receive display output and use etc but as soon as PC is under load the display is lost and PC remains turned on but soon restarts itself. Sometimes I can run heaven benchmark with no issues but as soon as I load up a game (warzone) it crashes.

 

I hope the RMA will fix the problem but still welcome to other suggestions.

[namireads]

Update:

Corsair RMA process was great and they sent me a new replacement PSU no questions asked! I've since installed the PSU and guess what... problem still persists

 

Weirdly, this time I don't get the POST 97 error and instead the display goes off and I have to manually restart the PC. I get to boot back into windows everytime however (before I would get stuck on post 97).

 

Still convinced this is a power draw problem and it only occurs under GPU load and having already replaced the GPU, I tried undervolting my card with MSI afterburner and using Heaven Benchmark to test:

 

Before Undervolt (stock):

GPU temp: 53^oC

GPU MHz: 1990 - 2010 MHz

GPU power draw: 247W

Mem: 1526MB

Mem MHz: 9502 MHz

CPU temp: 43^oC

CPU power draw: 76W

 

After Undervolt:

GPU temp: 54^oC

GPU MHz: 1840 - 1850MHz

GPU power draw: 2215W

Mem: 1380MB

Mem MHz: 9502 MHz

CPU temp: 51^oC

CPU power draw: 120W

 

So far after running heaven for 20 mins and playing Warzone for 1 hour the issue hasn't returned. I have also considerably dropped my graphics quality in-game and experience about a 30-40% decrease in FPS. Heaven was running in Extreme set-up when stress testing.

 

Was my card running too high as 'stock' (It's never been OC'ed by myself) but the switch on the GPU is set to "BIOS OC" as opposed to the "Silent" setting?

 

I'll welcome any thoughts as to the issue here...

[-rascal-]

The RTX 3070Ti, with high transient power spikes, might be causing the PSU to trip, causing the system to shut down.

 

If the GPU has a peak power draw of ~247W power draw.

Assuming a 2.5x transient = you are looking at ~617W for the GPU alone.

Add another up to 200W for the CPU + motherboard + RAM , etc,

When that spike happens, it will probably trip the OPP or OCP on the PSU.

 

Your symptoms sounds like what is being reported / investigated here by 'Tech Gezus'.

 

 

[SamClan]

13 hours ago, namireads said:

I'll welcome any thoughts as to the issue here...

set it to silent and check if the problem is still there. 

 

to be honest I thing it is a Motherboard problem, you said you testet the GPU in other system and it worked fine there. 

 

have you Resizable bar turned on? you can check with GPU-Z 

 

 

why not just put the GPU in another PCIE 3.0 x16 slot? and test if it works 

[namireads]

16 hours ago, -rascal- said:

The RTX 3070Ti, with high transient power spikes, might be causing the PSU to trip, causing the system to shut down.

 

If the GPU has a peak power draw of ~247W power draw.

Assuming a 2.5x transient = you are looking at ~617W for the GPU alone.

Add another up to 200W for the CPU + motherboard + RAM , etc,

When that spike happens, it will probably trip the OPP or OCP on the PSU.

 

Your symptoms sounds like what is being reported / investigated here by 'Tech Gezus'.

 

 

Thanks for the insight and the video - I'm convinced this is the problem. Not really sure where to go from here apart from upgrading the PSU and take a hit on my current PSU. My problem is that my case will only fit SFF PSUs and good quality 1000w ones are limited and expensive

 

5 hours ago, SamClan said:

set it to silent and check if the problem is still there. 

 

to be honest I thing it is a Motherboard problem, you said you testet the GPU in other system and it worked fine there. 

 

have you Resizable bar turned on? you can check with GPU-Z 

 

 

why not just put the GPU in another PCIE 3.0 x16 slot? and test if it works 

I set it to silent and checked today, the problem immediately returned. Resizable bar is turned on and the problem persists in the other PCIE slot (although it wouldn't be as feasible to leave it here as in this slot it is literally is a couple mm away from the bottom case fans).

[namireads]

Update: Swapped PSU to Silverstone SX1000W. Whichy didn't make a difference except that I could run heaven for an extra 1.5mins before display output lost. Currently got GPU in second PCIE slot - this buys me another minute of heaven before display lost..

 

Contacted Asus to try RMA mobo and might try replace GPU while am it. Pretty much replacing whole PC at this stage. *sigh*

 

Here's a video of what happens: https://drive.google.com/file/d/1mCj244zJjBUqgtPgUUUNBR59tO3lnnGt/view?usp=sharing

 

Hopefully someone could spot diagnose the issue from the vid

[namireads]

Could the error be due to RAM incompatibility?

 

I run 2x8GB of Corsair Dominator Platinum RGB (CMT16GX4M2E3200C16)

 

I can't find it on the supported list here: https://rog.asus.com/motherboards/rog-strix/rog-strix-z590-e-gaming-wifi-model/helpdesk_qvl_memory/