F5F Stay Refreshed Hardware Desktop Games failing when using two GPUs might indicate a power supply issue.

Games failing when using two GPUs might indicate a power supply issue.

Games failing when using two GPUs might indicate a power supply issue.

L
LooseDawg
Senior Member
628
08-27-2024, 12:00 PM
#1
I've been troubleshooting crashes with two GPUs on my PC: an old RX 5700 and a brand new RTX 4070 Ti Super OC. Here’s the summary of the errors. I used BlueScreenView to extract the minidump files and this is what I found. RX 5700: Crashes with VIDEO_SCHEDULER_INTERNAL_ERROR (0x119), caused by watchdog.sys . RTX 4070 Ti: Crashes with DRIVER_IRQL_NOT_LESS_OR_EQUAL (0xD1), caused by nvlddmkm.sys . System Specs: Ryzen 7 7700, Gigabyte B650M Aorus Elite AX, Corsair 850W PSU, 32GB DDR5 RAM. Troubleshooting so far: Clean driver installs (DDU for both NVIDIA and AMD). Stress tests (no issues in OCCT/FurMark). No WHEA errors in Event Viewer. RMA'd the 4070 Ti (passed their tests). Tested with both GPUs – same issues. Could this be a PSU, motherboard, or software issue? Both logs suggest GPU-related problems, but it’s happening on two different GPUs. Any ideas? Also, the RX 5700 seems to be more stable. I can run Overwatch 2, for instance, for much longer without crashing compared to the 4070 which crashes within minutes. I know the 4070's tdp is WAY higher, so, could this indicate a faulty PSU? Here are the full logs of the errors extracted with BlueScreenView: RX 5700 ================================================== Dump File : 020525-15062-01.dmp Crash Time : 05/02/2025 10:51:20 Bug Check String : Bug Check Code : 0x00000119 Caused By Driver : watchdog.sys Parameter 1 : 00000000 00000001 Parameter 2 : 00000000 00000000 Parameter 3 : 00000000 00082737 Parameter 4 : ffffd10d c72c2000 Caused By Address : watchdog.sys+1316 File Description : Product Name : Company : File Version : Processor : x64 Crash Address : ntoskrnl.exe+4b85d0 Stack Address 1 : Stack Address 2 : Stack Address 3 : Computer Name : Full Path : C:\WINDOWS\Minidump\020525-15062-01.dmp Processors Count : 16 Major Version : 15 Minor Version : 26100 Dump File Size : 2.829.965 Dump File Time : 05/02/2025 10:51:53 ================================================== 4070 Ti Super OC ================================================== Dump File : 012925-9859-01.dmp Crash Time : 29/01/2025 20:37:33 Bug Check String : DRIVER_IRQL_NOT_LESS_OR_EQUAL Bug Check Code : 0x000000d1 Caused By Driver : nvlddmkm.sys Parameter 1 : ffffad8c b2d02f6c Parameter 2 : 00000000 00000007 Parameter 3 : 00000000 00000000 Parameter 4 : fffff801 8e5eb6f9 Caused By Address : nvlddmkm.sys+1b6f9 File Description : Product Name : Company : File Version : Processor : x64 Crash Address : ntoskrnl.exe+4b85d0 Stack Address 1 : Stack Address 2 : Stack Address 3 : Computer Name : Full Path : C:\WINDOWS\Minidump\012925-9859-01.dmp Processors Count : 16 Major Version : 15 Minor Version : 26100 Dump File Size : 3.883.527 Dump File Time : 29/01/2025 20:38:01 ==================================================
L
LooseDawg
08-27-2024, 12:00 PM #1

I've been troubleshooting crashes with two GPUs on my PC: an old RX 5700 and a brand new RTX 4070 Ti Super OC. Here’s the summary of the errors. I used BlueScreenView to extract the minidump files and this is what I found. RX 5700: Crashes with VIDEO_SCHEDULER_INTERNAL_ERROR (0x119), caused by watchdog.sys . RTX 4070 Ti: Crashes with DRIVER_IRQL_NOT_LESS_OR_EQUAL (0xD1), caused by nvlddmkm.sys . System Specs: Ryzen 7 7700, Gigabyte B650M Aorus Elite AX, Corsair 850W PSU, 32GB DDR5 RAM. Troubleshooting so far: Clean driver installs (DDU for both NVIDIA and AMD). Stress tests (no issues in OCCT/FurMark). No WHEA errors in Event Viewer. RMA'd the 4070 Ti (passed their tests). Tested with both GPUs – same issues. Could this be a PSU, motherboard, or software issue? Both logs suggest GPU-related problems, but it’s happening on two different GPUs. Any ideas? Also, the RX 5700 seems to be more stable. I can run Overwatch 2, for instance, for much longer without crashing compared to the 4070 which crashes within minutes. I know the 4070's tdp is WAY higher, so, could this indicate a faulty PSU? Here are the full logs of the errors extracted with BlueScreenView: RX 5700 ================================================== Dump File : 020525-15062-01.dmp Crash Time : 05/02/2025 10:51:20 Bug Check String : Bug Check Code : 0x00000119 Caused By Driver : watchdog.sys Parameter 1 : 00000000 00000001 Parameter 2 : 00000000 00000000 Parameter 3 : 00000000 00082737 Parameter 4 : ffffd10d c72c2000 Caused By Address : watchdog.sys+1316 File Description : Product Name : Company : File Version : Processor : x64 Crash Address : ntoskrnl.exe+4b85d0 Stack Address 1 : Stack Address 2 : Stack Address 3 : Computer Name : Full Path : C:\WINDOWS\Minidump\020525-15062-01.dmp Processors Count : 16 Major Version : 15 Minor Version : 26100 Dump File Size : 2.829.965 Dump File Time : 05/02/2025 10:51:53 ================================================== 4070 Ti Super OC ================================================== Dump File : 012925-9859-01.dmp Crash Time : 29/01/2025 20:37:33 Bug Check String : DRIVER_IRQL_NOT_LESS_OR_EQUAL Bug Check Code : 0x000000d1 Caused By Driver : nvlddmkm.sys Parameter 1 : ffffad8c b2d02f6c Parameter 2 : 00000000 00000007 Parameter 3 : 00000000 00000000 Parameter 4 : fffff801 8e5eb6f9 Caused By Address : nvlddmkm.sys+1b6f9 File Description : Product Name : Company : File Version : Processor : x64 Crash Address : ntoskrnl.exe+4b85d0 Stack Address 1 : Stack Address 2 : Stack Address 3 : Computer Name : Full Path : C:\WINDOWS\Minidump\012925-9859-01.dmp Processors Count : 16 Major Version : 15 Minor Version : 26100 Dump File Size : 3.883.527 Dump File Time : 29/01/2025 20:38:01 ==================================================

M
Mai_Games
Member
105
08-27-2024, 01:54 PM
#2
Verify PSU voltages using HWinfo Test CPU and memory in OCCT. Check the event viewer for any indications about the crash.
M
Mai_Games
08-27-2024, 01:54 PM #2

Verify PSU voltages using HWinfo Test CPU and memory in OCCT. Check the event viewer for any indications about the crash.

P
PokStor
Junior Member
13
08-28-2024, 03:22 AM
#3
Voltages seem fine although I'm not an expert in PC voltages. I've read that all rails must stay within 5% of their rated voltages. They seem very stable in this sense. CPU is also fine in the tests, temperatures are great as well. Here is the log the last crash in event viewer: The computer has rebooted from a bugcheck. The bugcheck was: 0x00000119 (0x0000000000000001, 0x0000000000000000, 0x0000000000082737, 0xffffd10dc72c2000). A dump was saved in: C:\WINDOWS\Minidump\020525-15062-01.dmp. Report Id: 6e5e5028-5606-42a9-9b4e-4b973bf95d82. In the minidump file I found out it was a watchdog.sys . driver error that triggered the crash. My PC did a hard reset when Diablo 4 crashed last time.
P
PokStor
08-28-2024, 03:22 AM #3

Voltages seem fine although I'm not an expert in PC voltages. I've read that all rails must stay within 5% of their rated voltages. They seem very stable in this sense. CPU is also fine in the tests, temperatures are great as well. Here is the log the last crash in event viewer: The computer has rebooted from a bugcheck. The bugcheck was: 0x00000119 (0x0000000000000001, 0x0000000000000000, 0x0000000000082737, 0xffffd10dc72c2000). A dump was saved in: C:\WINDOWS\Minidump\020525-15062-01.dmp. Report Id: 6e5e5028-5606-42a9-9b4e-4b973bf95d82. In the minidump file I found out it was a watchdog.sys . driver error that triggered the crash. My PC did a hard reset when Diablo 4 crashed last time.

H
hunchmuffin6
Member
209
08-28-2024, 04:07 AM
#4
I've seen those BSOD alerts before—like when I pushed my graphics card too hard (watchdog error), with a failing i7 using the irql driver, or when I tried to overclock an i7. It might be connected to memory issues, but you should double-check that first since my memory of it is fuzzy.
H
hunchmuffin6
08-28-2024, 04:07 AM #4

I've seen those BSOD alerts before—like when I pushed my graphics card too hard (watchdog error), with a failing i7 using the irql driver, or when I tried to overclock an i7. It might be connected to memory issues, but you should double-check that first since my memory of it is fuzzy.

R
riptide1680
Member
169
08-28-2024, 05:10 AM
#5
You're referring to RAM, and you're preparing to run the Memtest86+ at least four times to verify everything is working correctly.
R
riptide1680
08-28-2024, 05:10 AM #5

You're referring to RAM, and you're preparing to run the Memtest86+ at least four times to verify everything is working correctly.

D
DannyBriere
Member
53
08-28-2024, 07:54 AM
#6
bump
D
DannyBriere
08-28-2024, 07:54 AM #6

bump