Game Causing PC Hard Crash

Hey guys!

New to the forum and DCS World but having a bit of a problem. Everytime I get into a DCS World Flight, I crash within about a minute of flying if I last that long. This is happening in both steam edition, and the standalone. No mods or anything extra installed, just the instant flights. My GPU fans spin up, the light on the gpu turns off, and my screens go black. Sound continues. I got the PC through NZXT BLD and I've sent it back to them. They swapped the mobo, CPU, PSU, and GPU and they think it may be some software issue.

I've tried reinstalling all drivers, flashed the bios, wiped and did a clean install of windows, and turned off almost all background apps. I run DCS on low graphics, and already ensured the power management settings are correct for the program on the NVIDIA control panel. I also monitor all temps and nothing seems to be spiking before the crash. I'm totally at a loss. This is a new PC and it should not be crashing.

Reliability Report:

Problem Event Name: LiveKernelEvent

Code: 141

Parameter 1: ffffa205265b4010

Parameter 2: fffff80444d8b9b4

Parameter 3: 0

Parameter 4: 1f44

OS version: 10_0_18363

Service Pack: 0_0

Product: 768_1

OS Version: 10.0.18363.2.0.0.768.101

Locale ID: 1033

Bucket ID: LKD_0x141_Tdr:6_IMAGE_nvlddmkm.sys_Unknown

Server information: 6067a0be-6592-41c2-a17e-cb6ff3c2d85b

PC Specs:

CPU: i7-9700K

CPU Cooler: Kraken X62

GPU: RTX 2070 Super Zotac Gaming Twin Fan

Ram: 2x G.Skill F4-3200C16-8GTZR DDR4

Mobo: MPG Z390 Gaming Pro Carbon

Storage: 1TB Intel 660p SSD, 2TB HDD

PSU: NZXT E850 Gold Digital PSU

Let me know if there is any other information I should provide!

Log:

dcslog.txt.txt

Last line of the log says:

ERROR DX11BACKEND: DX device removed. Reason: 0x887A0005

If you search for the error number (which is a Windows one) you get this:

0x887A0005 - DXGI_ERROR_DEVICE_REMOVED The video card has been physically removed from the system, or a driver upgrade for the video card has occurred. The application should destroy and recreate the device. For help debugging the problem, call ID3D10Device::GetDeviceRemovedReason.

If you search for this error you find a lot of them for all sorts of games (not DCS only).

One thing pops to mind: does your videocard receives enough power? Once it spools up when the game starts to run, it will draw more power from the powersupply. Looking at your specs the powersupply should be enough so you might take a look at the connections. Also, what happens if you run a stress program for your videocard?

Ran stress tests, nothing happen. PSU was RMA'd and replaced. I replaced the PCI cable as well and made sure its in properly. New GPU, New MOBO, same issue. Must be a problem with the installation I'm thinking.

I ran stress tests and no crashing. I have great cooling (and my rooms pretty cold too) so the temps never got above 140 Fahrenheit. Also never hit a point that i didn't have enough power.

Also no overclocking on anything. Running low graphics. Shouldn't cause much stress at all. Is there a way to also limit my frames in the game?

Ran stress tests, nothing happen. PSU was RMA'd and replaced. I replaced the PCI cable as well and made sure its in properly. New GPU, New MOBO, same issue. Must be a problem with the installation I'm thinking.

 

I ran stress tests and no crashing. I have great cooling (and my rooms pretty cold too) so the temps never got above 140 Fahrenheit. Also never hit a point that i didn't have enough power.

 

Also no overclocking on anything. Running low graphics.

 

Shouldn't cause much stress at all.

 

Is there a way to also limit my frames in the game?

Things to try:

1. Set a fixed large Windows pagefile of 16gb or more as your DCS seems to show it as too low.
   
2. If you want to use frame limiting, nvidia has just added that feature to game profiles in latest drivers, otherwise you can do it with MSI Afterburner and rivatuner.
   
3. Your issue sounds similar to an issue I had when I bought a gigabyte gtx-780 years ago, it did the same as your describing and only with DCS, all other games worked fine. (Problem I found was that the GPU voltage was too low to function correctly at low graphic loads, I remaped the GPU BIOS to ramp up the voltage slightly under different loads / MHZ from 0.888v to no less than 1.1v and it worked fine for 4 years after that), rather extreme, but gigabyte was not use in helping resolve the issue.
   
4. Try using GPU-Z to monitor GPU load and voltage, noting operating voltages at different graphic loads and see if you can spot a point that it fails in game.
   
5. Also you using HDMI or Display Port connection to monitors, DP particularly have some faulty cables out there supplying voltage on pin 20 (from memory) which they shouldn't, this can cause strange issues and kill graphics card.
   

Best of luck with it, I know how very frustrating it is.

Regards, Ian.

Take a look here at the TDR issue, it may help

 

https://forums.eagle.ru/showthread.php?t=232712

I already gave it a look but there was no fix. I thought this one was applicable. Was there a solid fix to it I missed?

Things to try:

 

1. Set a fixed large Windows pagefile of 16gb or more as your DCS seems to show it as too low.
   
2. If you want to use frame limiting, nvidia has just added that feature to game profiles in latest drivers, otherwise you can do it with MSI Afterburner and rivatuner.
   
3. Your issue sounds similar to an issue I had when I bought a gigabyte gtx-780 years ago, it did the same as your describing and only with DCS, all other games worked fine. (Problem I found was that the GPU voltage was too low to function correctly at low graphic loads, I remaped the GPU BIOS to ramp up the voltage slightly under different loads / MHZ from 0.888v to no less than 1.1v and it worked fine for 4 years after that), rather extreme, but gigabyte was not use in helping resolve the issue.
   
4. Try using GPU-Z to monitor GPU load and voltage, noting operating voltages at different graphic loads and see if you can spot a point that it fails in game.
   
5. Also you using HDMI or Display Port connection to monitors, DP particularly have some faulty cables out there supplying voltage on pin 20 (from memory) which they shouldn't, this can cause strange issues and kill graphics card.
   

Best of luck with it, I know how very frustrating it is.

 

Regards, Ian.

I had always heard DP is better, but I can try out the HDMI and see if ti makes a difference. I have two monitors thoguh and one uses HDMI, but both turn off when my GPU crashes either way. I'll reinstall GPU-Z and post the log here if it crashes. Then I can tryout some voltage modding. Thanks

Just increasing the time out value, it helps some users on the newer cards.

I agree. I want to try this. How do I change it? Didn't see it in that forum or online. Do I need NVIDIA NSights? Or does NVIDIA Control panel or windows handle it.

There is a microsoft article here explaining it may help

 

https://docs.microsoft.com/en-gb/windows-hardware/drivers/display/timeout-detection-and-recovery

Unfortunately, didn't work when upping to 8 seconds. Even ran the game 1920x1080 with a very limited FPS instead of 2560x1440. Made no difference.

See Crash log from GPU-Z attached. Crash occurs at 2020-01-24 12:03:16.782. Next thought is to replaced my GPU thermal paste.

GPU-Z Sensor Log_3.1.txt

I guess it happens a bit before that at: 2020-01-24 12:03:14.160

Just compare the one where the values are still OK: 2020-01-24 12:03:12.454

with the one where it doesn't: 2020-01-24 12:03:14.160

But there is some other stuff to be seen at that time:

Date: 2020-01-24 12:03:12.454 --------- 2020-01-24 12:03:14.160

GPU Clock [MHz]: 1980.0 ---------------- 1980.0 --> OK

Memory Clock [MHz]: 1750.2 -------------1750.2 --> OK

GPU Temperature [°C]: 40.0 ------------- 0.0 --> NOK

Fan 1 Speed (%): 31 -------------------- 31 --> OK

Fan 1 Speed (RPM): 1081 --------------- 1081 --> OK

Fan 2 Speed (%): 32 ------------------- 32 --> OK

Fan 2 Speed (RPM): 1101 --------------- 1101 --> OK

Memory Used [MB]: 2889 --------------- 2889 --> OK

GPU Load [%]: 46 ---------------------- 46 --> OK

Memory Controller Load [%]: 15 --------- 15 --> OK

Video Engine Load [%]: 0 --------------- 0 --> OK

Bus Interface Load [%]: 1 -------------- 1 --> OK

Power Consumption (W): 101,7 --------- 0.0 --> NOK

Power Consumption (%): 47,3 ---------- 0.0 --> NOK

PerfCap Reason []: 16 ------------------ 0 --> NOK (no idea what PrefCap stands for)

VDDC [V]: 1.0370 ---------------------- 0.0000 --> NOK

CPU Temperature [°C]: 32.0 -------------34.0 --> OK

System Memory Used [MB]: 9594 ------- 9596 --> OK

At that moment the GPU temp goes from 40°c to 0°c all of a sudden, that's not normal (has nothing to do with thermal paste).

The Power Consumption goes to 0 and also the VDDC (a voltage setting) drops to zero. These last two worry me a bit.

All connection from GPU-Z to the card doesn't seem to be lost because it still reads other stuff fine.

I know. I just ran more stress tests including First Strike and the other DX12 one and I pass all the pre-crash thresholds shown in here without trouble. I can play other games that don't cause these problems at all but are far more intensive.

Any other info I can provide? Or any ideas haha.

I guess it happens a bit before that at: 2020-01-24 12:03:14.160

Just compare the one where the values are still OK: 2020-01-24 12:03:12.454

with the one where it doesn't: 2020-01-24 12:03:14.160

 

But there is some other stuff to be seen at that time:

Date: 2020-01-24 12:03:12.454 --------- 2020-01-24 12:03:14.160

GPU Clock [MHz]: 1980.0 ---------------- 1980.0 --> OK

 

Memory Clock [MHz]: 1750.2 -------------1750.2 --> OK

 

GPU Temperature [°C]: 40.0 ------------- 0.0 --> NOK

 

Fan 1 Speed (%): 31 -------------------- 31 --> OK

 

Fan 1 Speed (RPM): 1081 --------------- 1081 --> OK

 

Fan 2 Speed (%): 32 ------------------- 32 --> OK

 

Fan 2 Speed (RPM): 1101 --------------- 1101 --> OK

 

Memory Used [MB]: 2889 --------------- 2889 --> OK

 

GPU Load [%]: 46 ---------------------- 46 --> OK

 

Memory Controller Load [%]: 15 --------- 15 --> OK

 

Video Engine Load [%]: 0 --------------- 0 --> OK

 

Bus Interface Load [%]: 1 -------------- 1 --> OK

 

Power Consumption (W): 101,7 --------- 0.0 --> NOK

Power Consumption (%): 47,3 ---------- 0.0 --> NOK

 

PerfCap Reason []: 16 ------------------ 0 --> NOK (no idea what PrefCap stands for)

VDDC [V]: 1.0370 ---------------------- 0.0000 --> NOK

 

CPU Temperature [°C]: 32.0 -------------34.0 --> OK

 

System Memory Used [MB]: 9594 ------- 9596 --> OK

 

 

At that moment the GPU temp goes from 40°c to 0°c all of a sudden, that's not normal (has nothing to do with thermal paste).

The Power Consumption goes to 0 and also the VDDC (a voltage setting) drops to zero. These last two worry me a bit.

All connection from GPU-Z to the card doesn't seem to be lost because it still reads other stuff fine.

Check all your power supply connections? on the board also? make sure everything is seated and firmly attached?

make sure the ram is seated correctly.

sounds odd. maybe PSU cords power cable to the Video card?

PCIE slot the card is in seated?

bad CPU placement socket one of the pins is bent?

I had always heard DP is better, but I can try out the HDMI and see if it makes a difference. I have two monitors though and one uses HDMI, but both turn off when my GPU crashes either way.

Yes, DP is definitely better, but I was only suggesting testing with HDMI and check that DP cable isn't one of the faulty ones with power fault on pin 20 that allows power from the monitor back into the graphics card (Should not do this), even when your system is turned off.

https://www.tenforums.com/graphic-cards/137524-experienced-pin-20-display-port-problem.html

As for that GPU-Z report, wow, that is astounding, not sure on how to deal with those issues.

1. Post a DXDiag report, might give us some clues.
   
2. Have a look through "Windows error reports" in event viewer.
   
3. What type of PCIE power connectors are going to the graphics card, 8+6 pin or 8+8 pin, are they from the PSU on separate power feed or some other arrangement (Off one power feed with a splitter to two feeds).
   

Regards, Ian.

Yes, DP is definitely better, but I was only suggesting testing with HDMI and check that DP cable isn't one of the faulty ones with power fault on pin 20 that allows power from the monitor back into the graphics card (Should not do this), even when your system is turned off.

 

https://www.tenforums.com/graphic-cards/137524-experienced-pin-20-display-port-problem.html

 

As for that GPU-Z report, wow, that is astounding, not sure on how to deal with those issues.

 

1. Post a DXDiag report, might give us some clues.
   
2. Have a look through "Windows error reports" in event viewer.
   
3. What type of PCIE power connectors are going to the graphics card, 8+6 pin or 8+8 pin, are they from the PSU on separate power feed or some other arrangement (Off one power feed with a splitter to two feeds).
   

Regards, Ian.

1. What would help in the DXDIAG report? And how do I get that? Through the DXDIAG.exe?

2. Pasting some odd stuff I found in the event viewer below.

3. I use an 8+6 pin. Same cord coming from the PSU. I can try replacing that cable next.

Adding the summary of windows events. Haven't tried DCS the past 2 days because I hate seeing my computer crash like this lol

Hey guys!

 

New to the forum and DCS World but having a bit of a problem. Everytime I get into a DCS World Flight, I crash within about a minute of flying if I last that long. This is happening in both steam edition, and the standalone. No mods or anything extra installed, just the instant flights. My GPU fans spin up, the light on the gpu turns off, and my screens go black. Sound continues. I got the PC through NZXT BLD and I've sent it back to them. They swapped the mobo, CPU, PSU, and GPU and they think it may be some software issue.

 

I've tried reinstalling all drivers, flashed the bios, wiped and did a clean install of windows, and turned off almost all background apps. I run DCS on low graphics, and already ensured the power management settings are correct for the program on the NVIDIA control panel. I also monitor all temps and nothing seems to be spiking before the crash. I'm totally at a loss. This is a new PC and it should not be crashing.

 

Reliability Report:

Problem Event Name: LiveKernelEvent

Code: 141

Parameter 1: ffffa205265b4010

Parameter 2: fffff80444d8b9b4

Parameter 3: 0

Parameter 4: 1f44

OS version: 10_0_18363

Service Pack: 0_0

Product: 768_1

OS Version: 10.0.18363.2.0.0.768.101

Locale ID: 1033

Bucket ID: LKD_0x141_Tdr:6_IMAGE_nvlddmkm.sys_Unknown

Server information: 6067a0be-6592-41c2-a17e-cb6ff3c2d85b

 

 

PC Specs:

CPU: i7-9700K

CPU Cooler: Kraken X62

GPU: RTX 2070 Super Zotac Gaming Twin Fan

Ram: 2x G.Skill F4-3200C16-8GTZR DDR4

Mobo: MPG Z390 Gaming Pro Carbon

Storage: 1TB Intel 660p SSD, 2TB HDD

PSU: NZXT E850 Gold Digital PSU

 

Let me know if there is any other information I should provide!

 

Log:

I had a similar problem. But it died w/in minutes of playing PUBG. And only PUBG. But since software can't turn off HW, I suspected my power supply. But it took me a long time to get there because I did all the stress tests out there. Cinebench, memtest, you name it.

After I replaced my PSU all my problems magically disappeared.

Also, there was another instance where one of my four memory sticks went bad. That was more hit and miss, but I was finally able to nail it down to won specific memory bank.

So I would try using a different cable for the power to your GPU. Make sure it's seated and the contacts are cleaned. Check for bios updates. Make sure you're no OC'ing the CPU or GPU as a start. And check what speed the memory is set to in BIOS.

Finally, use GPU-Z or MSIs of the world to check for thermal issues on the CPU/GPU or even memory.

OP:

You would need to get more detailed with your steps going forward because now it's more complicated as I can't figure out any clue anymore. If you RMAd like almost everything, ...

Did you build/assemble the computer or they did? Or who did?

Does this happen with any other game?

When you DXDIAG.exe there's a button Save All Information, and there you go, but zipp the file for transfer it makes it much smaller, also there's some private info in there you may want to remove like GUIDs/HWIDs/Usernames/Real names etc, not for us, but for the internetz.

You can't replace PSU cables just like that, if they're modular they're proprietary, and unlikely the power cable is the problem, it would haven't even boot if that were the case.

1. What would help in the DXDIAG report? And how do I get that? Through the DXDIAG.exe?

 

2. Pasting some odd stuff I found in the event viewer below.

 

3. I use an 8+6 pin. Same cord coming from the PSU. I can try replacing that cable next.

1. Dxdiag = Win + R keys -> Dxdiag.exe -> Save All information - > Zip file and Post.
   
2. Event viewer = Nothing relevant to problems.
   
3. GPU Power = Should have two individual cables going to GPU from the PSU -> PCIe-1 and PCIe-2 or PCI-03.
   

Had a look at the specs of the PSU:

1. They look adequate, but I suspect it anyway (Over power protection kicking in maybe).
   
2. I also noticed that NZXT have monitoring software for the PSU (NZXT CAM), install and use that to monitor the Wattage and Amperage loads of the 12v rails particalery when testing to see if there are any overloads or anonomolies - https://www.nzxt.com/products/e850.
   
3. It is mentioned that the 12v rails can be changed via software to split the amperage to each 12v rail/s, ie 70a single rail or 25a/25a/20a, you may want to experiment with these settings.
   
4. It is also mentioned that the overload protection threshold is programmable, may need increasing or experimented with.
   

Regards, Ian.

This was built by NZXT BLD and I don't 100% believe that they actually swapped things out like they said. But I found a loose screw in it too (turned out to be for an LED strip) so don't know what to think. This is actually the only game which is ridiculous.

I do use NZXT CAM. All looks fine but maybe I should turn down the default amps? Its at 60 at the moment for both pins by default.

Thanks for all the help everyone :) I'll update you if I find anything that helps fix this tomorrow.

i had this same error and pc would crash/freeze and id have to restart the pc . i threw a different psu in it and flew for a few hours without issues. sucks because a week ago my good motherboard got fried and i couldnt figure it out as the old psu seems to work perfectly fine. just lucky it didnt fry the cheap backup. ill keep updated if i get issues again.

PSUs can be sneaky, it's very easy to get confused with rails and amps, Wattage doesn't really mean that much as it's being portrayed.

I was absolutely capping the +12V line in terms of amps and didn't even knew it as the previous PSU was just that good and kept chugging along, that's why I was getting weird behavior in other cases (like weird drops of screen going black but PC still running, freezes, BSODs), even tho all games worked and I played them fine.

So now I got a proper single-rail PSU, actually this model has a physical switch to make it single or multi-rail, but I will never by multi-rail only PSUs again if I can, it takes planning and it takes a well calculated approach, which I don't want to be bothered with as I change components and modify my system on almost a tri monhly basis as I maintain it, upgrade HDDs, change stuff around, more HDDs, changing this, that, etc. Multi-Rail would be fine for like OEM when the PC is under warranty and nothing will be changed from the specs in the beginning for that period.

This was built by NZXT BLD and I don't 100% believe that they actually swapped things out like they said. But I found a loose screw in it too (turned out to be for an LED strip) so don't know what to think. This is actually the only game which is ridiculous.

 

 

I do use NZXT CAM. All looks fine but maybe I should turn down the default amps? Its at 60 at the moment for both pins by default.

 

Thanks for all the help everyone :) I'll update you if I find anything that helps fix this tomorrow.

I said the same thing about PUBG. "it's the only game that does it, it must be PUBG" But you're getting BSOD, where I was getting power off.

How many memory sticks do you have? I'd start taking them out (in pairs or singly depending on your layout) to isolate a bad memory chip. Running MEMTEST for 24 hours didn't catch it in my case. grr.

FYI // Maybe OT::

DP Pin20 Power issue can cause hardware failure and damage according to Dell Technical Support.

I am myself troubleshooting such a situation with a brand new Dell combo, all came from Dell, PC, Monitor 4k DP, Cable came with Monitor, keyboard mouse, everything.

The Dell guy comes on Monday visiting our customer, replacing the mobo and PSU to exclude possible damage.

Yes, we changed the DP cable to one that Dell has sent that is said to not have that Pin active.

If that is your source of problem, good luck.

* In our case, the PC turn ON but come to a halt after about 2-5 seconds, Before the Dell Logo shows up, far before the OS gets involved. If we turn the PC on with the monitor turned OFF and wait to turn it on until the Logo shows it's all well. Dell is trying to solve it for a few months now with no success. Now we will see if new parts do the trick.