New Build Restarting When Gaming - RTX 3080 / 5900X

[DeepC_]

Hi guys,

 

I built a new rig the other day, specs below, and it's randomly restarted on me 3 times now when gaming. It's happened once while playing Modern Warfare (was fine for several games then restarted randomly), and twice in a row when trying to boot up the Outriders Demo, specifically while the intro cinematic is playing. Not sure if this info is helpful but I can successfully leave my computer mining crypto (using both GPU and CPU) at full load and the temps don't go above 70° and it doesn't restart. Also, the first 2 days after building my rig there was a weird smell coming from the PC. I put this down to a 'new pc smell' but thought it's worth mentioning. I'm on day 3 after building the rig now and the smell is definitely lessening now.

 

The RAM and SSD (Data) was taken directly from my previous machine and had no issues. Also worth noting that I haven't overclocked anything, but I have enabled D.O.C.P on my motherboard in order to utilize my RAM correctly (No manual tweaking, just the automatic D.O.C.P settings). PC temps usually are around 20-40° idle depending on how hot my room is, and never go above 70° when gaming/full load.

 

Mobo - Asus Prime Z570-Pro

PSU - Corsair HX1000 1000W

GPU - Asus RTX 3080 STRIX OC White

CPU - AMD Ryzen 9 5900X

RAM - Corsair Vengeance Pro RGB White - 32gb 4x8gb

Cooler - NXZT Kraken Z3

SSD (OS) - Samsung 970 Evo Plus - Windows 10 Home 64 bit

SSD (Data) - Samsung 870 Evo

 

I did open up event viewer (although I don't fully understand what I'm looking at) and I've noticed the below extract that seems to be appearing each time it crashes.

 

Spoiler

 

Spoiler

-

System

- Provider       [ Name] Microsoft-Windows-Kernel-Power       [ Guid] {331c3b3a-2005-44c2-ac5e-77220c37d6b4}

EventID 41

Version 8

Level 1

Task 63

Opcode 0

Keywords 0x8000400000000002

- TimeCreated       [ SystemTime] 2021-04-08T16:20:16.5700616Z

EventRecordID 1441

Correlation

- Execution       [ ProcessID] 4       [ ThreadID] 8

Channel System

Computer DESKTOP-MC01ES4

- Security       [ UserID] S-1-5-18

-

EventData

BugcheckCode

0

BugcheckParameter1

0x0

BugcheckParameter2

0x0

BugcheckParameter3

0x0

BugcheckParameter4

0x0

SleepInProgress

0

PowerButtonTimestamp

0

BootAppStatus

0

Checkpoint

0

ConnectedStandbyInProgress

false

SystemSleepTransitionsToOn

0

CsEntryScenarioInstanceId

0

BugcheckInfoFromEFI

false

CheckpointStatus

0

CsEntryScenarioInstanceIdV2

0

LongPowerButtonPressDetected

false

 

I'm really not sure how to start troubleshooting here so I would be ever so grateful for some help!

 

Thanks,

[SignatureSigner]

Seems to be related to power issues... Maybe the power supply can't handle the spikes that the 30 series throws around. Doubtful but possible.

[Jacques247]

15 minutes ago, DeepC_ said:

Hi guys,

 

I built a new rig the other day, specs below, and it's randomly restarted on me 3 times now when gaming. It's happened once while playing Modern Warfare (was fine for several games then restarted randomly), and twice in a row when trying to boot up the Outriders Demo, specifically while the intro cinematic is playing. Not sure if this info is helpful but I can successfully leave my computer mining crypto (using both GPU and CPU) at full load and the temps don't go above 70° and it doesn't restart. Also, the first 2 days after building my rig there was a weird smell coming from the PC. I put this down to a 'new pc smell' but thought it's worth mentioning. I'm on day 3 after building the rig now and the smell is definitely lessening now.

 

The RAM and SSD (Data) was taken directly from my previous machine and had no issues. Also worth noting that I haven't overclocked anything, but I have enabled D.O.C.P on my motherboard in order to utilize my RAM correctly (No manual tweaking, just the automatic D.O.C.P settings). PC temps usually are around 20-40° idle depending on how hot my room is, and never go above 70° when gaming/full load.

 

Mobo - Asus Prime Z570-Pro

PSU - Corsair HX1000 1000W

GPU - Asus RTX 3080 STRIX OC White

CPU - AMD Ryzen 9 5900X

RAM - Corsair Vengeance Pro RGB White - 32gb 4x8gb

Cooler - NXZT Kraken Z3

SSD (OS) - Samsung 970 Evo Plus - Windows 10 Home 64 bit

SSD (Data) - Samsung 870 Evo

 

I did open up event viewer (although I don't fully understand what I'm looking at) and I've noticed the below extract that seems to be appearing each time it crashes.

 

  Reveal hidden contents

 

  Reveal hidden contents

-

System

- Provider       [ Name] Microsoft-Windows-Kernel-Power       [ Guid] {331c3b3a-2005-44c2-ac5e-77220c37d6b4}

EventID 41

Version 8

Level 1

Task 63

Opcode 0

Keywords 0x8000400000000002

- TimeCreated       [ SystemTime] 2021-04-08T16:20:16.5700616Z

EventRecordID 1441

Correlation

- Execution       [ ProcessID] 4       [ ThreadID] 8

Channel System

Computer DESKTOP-MC01ES4

- Security       [ UserID] S-1-5-18

-

EventData

BugcheckCode

0

BugcheckParameter1

0x0

BugcheckParameter2

0x0

BugcheckParameter3

0x0

BugcheckParameter4

0x0

SleepInProgress

0

PowerButtonTimestamp

0

BootAppStatus

0

Checkpoint

0

ConnectedStandbyInProgress

false

SystemSleepTransitionsToOn

0

CsEntryScenarioInstanceId

0

BugcheckInfoFromEFI

false

CheckpointStatus

0

CsEntryScenarioInstanceIdV2

0

LongPowerButtonPressDetected

false

 

I'm really not sure how to start troubleshooting here so I would be ever so grateful for some help!

 

Thanks,

Looks like a PSU issue, whats the rating on your psu? And the 3080 does have an issue with suddenly spiking way above the estimated wattage that it’s supposed to use.

[DeepC_]

6 minutes ago, SignatureSigner said:

Seems to be related to power issues... Maybe the power supply can't handle the spikes that the 30 series throws around. Doubtful but possible.

Thanks so much for the quick reply! 

 

I doubt it would be a PSU related issue since it's a 1000W PSU so should definitely be able to handle the spikes.

 

The only PSU related thing that I think could possibly be the cause would be that I am using the PCI-E cables that came with the PSU, the ones that split into two, so technically I am using 2 of the PSU PCI-E slots to cover all 3 8 pin connectors on the GPU.

 

Is this something that could cause an issue?

[DeepC_]

1 minute ago, Jacques247 said:

Looks like a PSU issue, whats the rating on your psu? And the 3080 does have an issue with suddenly spiking way above the estimated wattage that it’s supposed to use.

It's rated platinum. Could it have something to do with me using only two of the PCI-E PSU ports to cover all 3 8 pin connectors on the GPU, using the stock cables that came with the PSU? (As they split into two 8 pin connectors each)

[Jacques247]

1 minute ago, DeepC_ said:

It's rated platinum. Could it have something to do with me using only two of the PCI-E PSU ports to cover all 3 8 pin connectors on the GPU, using the stock cables that came with the PSU? (As they split into two 8 pin connectors each)

If the gpu asks for 3 8pins, then thats probably it. Not a professional but logic tells me that you are underpowering the gpu and that might be causing the pc to short circuit?

[DeepC_]

Just now, Jacques247 said:

If the gpu asks for 3 8pins, then thats probably it. Not a professional but logic tells me that you are underpowering the gpu and that might be causing the pc to short circuit?

I am still using 3 8 pins into the GPU, but two of the outputs into GPU are from 1 input port on PSU - like the below screenshot (although 8 pin)

 

Spoiler

 

[Jacques247]

Just now, DeepC_ said:

I am still using 3 8 pins into the GPU, but two of the outputs into GPU are from 1 input port on PSU - like the below screenshot (although 8 pin)

 

  Hide contents

 

Ohhhhhhhh, yeah, see if you can get 3 separate pci-e cables and then try....Maybe the psu aint sending enough power through the one pcie port

[CT854]

2 minutes ago, DeepC_ said:

*snip*

 

I could be mistaken, but I think the issue comes from the fact that you're using the same cable for two of the pins. The issue has to do with the current the cables themselves are rated for, so I've seen that it's highly encouraged to use three separate cables.  Someone check me on that, though.  It's certiainly worth a try if you have a third cable.

[Jacques247]

11 minutes ago, CT854 said:

 

I could be mistaken, but I think the issue comes from the fact that you're using the same cable for two of the pins. The issue has to do with the current the cables themselves are rated for, so I've seen that it's highly encouraged to use three separate cables.  Someone check me on that, though.  It's certiainly worth a try if you have a third cable.

My thought exactly.

[DeepC_]

6 minutes ago, Jacques247 said:

My thought exactly.

 

17 minutes ago, CT854 said:

 

I could be mistaken, but I think the issue comes from the fact that you're using the same cable for two of the pins. The issue has to do with the current the cables themselves are rated for, so I've seen that it's highly encouraged to use three separate cables.  Someone check me on that, though.  It's certiainly worth a try if you have a third cable.

Thanks both, I will swap out the cables and update the thread if the issue persists however as it's not extremely frequent not sure if I will know by end of the day!

 

Your prompt responses have been much appreciated.

[DeepC_]

On 4/8/2021 at 6:36 PM, CT854 said:

 

I could be mistaken, but I think the issue comes from the fact that you're using the same cable for two of the pins. The issue has to do with the current the cables themselves are rated for, so I've seen that it's highly encouraged to use three separate cables.  Someone check me on that, though.  It's certiainly worth a try if you have a third cable.

 

On 4/8/2021 at 6:47 PM, Jacques247 said:

My thought exactly.

Hi both, back again!

 

I have changed the cable over so now there are three separate PCI-E cables running into the GPU, booted up Outriders and the computer restarted before the splash cinematic began.

 

Have also now disabled D.O.C.P just in case it was that (as I was having boot issues because of D.O.C.P earlier), and have ran a CPU and GPU benchmark - both had no issues whatsoever.

 

Really not sure what else could be the issue! The RAM that I am using was taken straight from my other system and it was working perfectly on there so I am doubtful that it's faulty.

 

Any other ideas?

 

Thanks!

[KillzoneCX]

I have the same problem. Started with 850w Chieftec Bronze next i bought an HX1000 plat and i can go 80% power limit. Doom Eternal 30s in-game and black screen 100% of fan speed on GPU. RTX 3090 Aourus Master Ryzen 2700 16gb of DDR4 3Ghz Ram. I am 95% sure that is this a power supply issue. I will try to replace it with 1200w to check this if u maybe has some bigger PSU u can check it so I will buy same to stop searching the working PSU with 3090

 

EDIT: Yesterday i was able to make some research and tests. So I block the clock curve on 1815 MHz for core on 0.825v(max core clock in game will be 1815Mhz starting from 0.825v so even if gpu will be on 1v the clock will be 1815mhz base clock for me is 1785mhz something liek that) and restarts have stopped. I am thinking if that can be a bios issue with aggressive boost clocks. I Saw there 1.25v 2.2 GHz. Maybe this clock si so much unstable that makes shutdowns. Bu eventually test on 1200w Power supply will be best to approve theory about aggressive boost clocks on Aourus Master.

[DeepC_]

On 4/11/2021 at 7:21 PM, KillzoneCX said:

I have the same problem. Started with 850w Chieftec Bronze next i bought an HX1000 plat and i can go 80% power limit. Doom Eternal 30s in-game and black screen 100% of fan speed on GPU. RTX 3090 Aourus Master Ryzen 2700 16gb of DDR4 3Ghz Ram. I am 95% sure that is this a power supply issue. I will try to replace it with 1200w to check this if u maybe has some bigger PSU u can check it so I will buy same to stop searching the working PSU with 3090

 

EDIT: Yesterday i was able to make some research and tests. So I block the clock curve on 1815 MHz for core on 0.825v(max core clock in game will be 1815Mhz starting from 0.825v so even if gpu will be on 1v the clock will be 1815mhz base clock for me is 1785mhz something liek that) and restarts have stopped. I am thinking if that can be a bios issue with aggressive boost clocks. I Saw there 1.25v 2.2 GHz. Maybe this clock si so much unstable that makes shutdowns. Bu eventually test on 1200w Power supply will be best to approve theory about aggressive boost clocks on Aourus Master.

After performing a bios update on my motherboard I have since had 0 issues at all. It looks like my motherboard was flashed with a bios at the retailer I got it from in order to make it compatible with the 5900X however they didn't flash it with the most up to date version.

 

The only other things I have changed are disabling D.O.C.P and I am now using 3 8 pin PCI-E cables rather than 2 (with one daisy chained to fill 3 8 pin slots on the GPU) however I am pretty certain it was the bios update that resolved the issue.

 

Give that a try before purchasing a 1200W PSU because frankly a 750W should be fine for a 3080, to my knowledge.