RTX 2070 super gaming x trio crash/blackscreen -> hard reset

dopamin44

New member
PRIVATE E-2
Joined
Jun 8, 2020
Messages
7
hey guys, when i run the Kombustor stresstest, It runs for 50 sec and i get a black screen. The fans of my gpu are turning off, but my pc is still on. When i wait for a few minutes, the fans of my card ramp up to 100%, but nothing else happens. Only fix is to hard reset my system.

When i look in the event viewer, there is an Nvidia driver crash. but only when ive waited the few minutes until the gpu fans ramp to 100%. When i hard reset immediately after the black screen, my system speaker makes 1 long, 3 short beeps, meaning that no Graphic card was detected. I restart again and i boots, but nothing shows up in the event viewer.
After i DDU the driver and put it fresh on. I get no crashes in Kombustor, but only for some time. The crashes return, again around the 50 sec mark. When i DDU everything again, it will work again, until is doesnt.

Temperatures are all fine, GPU is around 60-68 ?C. I have not overclocked the card. I run this system for 6 months without problems, although i play with 60fps capped, so i never really stressed the GPU that much. i am clueless at this point, i just hope i dont have to RMA something. sorry for the bad english, its not my native language.^^

Gigabyte X570 Aorus Elite - Bios F20
Ryzen 3700x - stock
MSI RTS 2070 super gaming x trio - stock
Crucial Ballistix Sport 3200 - xmp active
be quiet pure power 11 600w
Windows 2004
Chipset Driver: 2.04.28.626
 

darkhawk

Well-known member
Global Moderator
Joined
Nov 20, 2012
Messages
12,849
I'll be perfectly honest, MSI and Gigabyte hardware really don't work well together. 

Hard to really say what the issue is, but it could be the driver. Maybe go back a few versions on the driver and see if that helps resolve the issues?
 

dopamin44

New member
PRIVATE E-2
Joined
Jun 8, 2020
Messages
7
well i tried from 441.86 till 455.38 every driver. the most stable is the 455.38, but this is a developer driver which is missing vulkan api. are there any reasons why msi and gigabyte hardware wont work well together?
 

darkhawk

Well-known member
Global Moderator
Joined
Nov 20, 2012
Messages
12,849
dopamin44 date=1592874413 said:
well i tried from 441.86 till 455.38 every driver. the most stable is the 455.38, but this is a developer driver which is missing vulkan api. are there any reasons why msi and gigabyte hardware wont work well together?
I can't give you a good technical reason. 
Generally, if they boot up, they work OK together.

But lately things haven't always gone that way, and I've been seeing more issues similar to your own.

As dumb as it sounds....are you sure the power supply is more than sufficient?
Looking at the specs for the card you have, it recommends a 650W+ power supply, and at 600W you're a bit under.
Also, for what it's worth, it's not a GREAT supply. It's OK, but I wouldn't push it to the limits (like this card is probably doing).
See : https://linustechtips.com/main/topic/1116640-psucultists-psu-tier-list/
Your supply is a Tier B+. Ok for a mid-range PC, but with a 2070 Super and a 3700X, I'm not sure it's suited for what you're trying to use it with.
Maybe consider a 700W+ unit?

The best I can suggest is to contact MSI directly via : >>How to contact MSI.<<
Maybe the card is bad, if you can't get things to run well together.
 

dopamin44

New member
PRIVATE E-2
Joined
Jun 8, 2020
Messages
7
The PSU was my first guess as well. But, the thing is, my system does not reboot on its own or something like that. The system is running, but only the gpu is crashing for some reason. I dont think i would get driver errors in event viewer if my PSU would be insufficent. Its just so weird
 

darkhawk

Well-known member
Global Moderator
Joined
Nov 20, 2012
Messages
12,849
Agreed. That's what one would normally think. But here's the thing. Your supply is OK, it's just not recommended for a hardcore system.

With the parts you have, you're in a middle ground. Your CPU is power hungry (they all are at that level honestly....), and add in a rather power hungry GPU and who knows what will happen.

Under a normal load, your CPU will draw 150W, not overclocked. Overclocked and you might be at 175W or so.
Add in chipset, motherboard, memory, any lights, etc....and the GPU, alone is another 235W. So between the GPU and CPU alone, you're already at just over 400W or so. 

Your supply is OK. If you don't have many drives, no lights, not many fans, etc.....you might be OK overall. At 400W for CPU and GPU, that does leave another 200W for the motherboard/memory/peripherals and such.......
However.....most likely, that supply can barely handle 600W is my guess. Maybe it can realistically only handle 500W, maybe 550W, for extended periods of time. Maybe less, if it's fully loaded during a game. 

Long story short, I'd use the link I previously suggested, and find a Tier A supply, if at all possible. Preferably one that is 700W.
 

dopamin44

New member
PRIVATE E-2
Joined
Jun 8, 2020
Messages
7
allright, i have ordered this PSU and test it out.
should be here on saturday.

but, can you tell me, why my system does not shut down/restart  when i get the blackscreen? when i hear of PSU issues, i always read about systems shutting down or restarting on its own. but thats not really happening for me. i have to hard reset my system manually when this happens. thats why i am still a bit sceptic. but hopefully this will fix the problem.
 

darkhawk

Well-known member
Global Moderator
Joined
Nov 20, 2012
Messages
12,849
The only reason I can really come up with, is that the supply isn't just dying or turning off, but the voltage output from the supply might be dropping a lot, and causing the card to not operate as expected, rather than just shutting off.

I would probably say you're at the edge of it working, and longer periods of activity cause the problem you're seeing.
 

dopamin44

New member
PRIVATE E-2
Joined
Jun 8, 2020
Messages
7
I see, lets hope you are right. but i would keep the new PSU anyway.^^
 

dopamin44

New member
PRIVATE E-2
Joined
Jun 8, 2020
Messages
7
it just happened again. so the psu wasnt the culprit. should i rma my card?
 

dopamin44

New member
PRIVATE E-2
Joined
Jun 8, 2020
Messages
7
here a dump from the crash
==================================================
Dump File : 080920-14812-01.dmp
Crash Time : 09.08.2020 13:13:39
Bug Check String : KMODE_EXCEPTION_NOT_HANDLED
Bug Check Code : 0x0000001e
Parameter 1 : ffffffff`c0000005
Parameter 2 : fffff801`23ac2c32
Parameter 3 : 00000000`00000000
Parameter 4 : ffffffff`ffffffff
Caused By Driver : nvlddmkm.sys
Caused By Address : nvlddmkm.sys+5e2c32
File Description :
Product Name :
Company :
File Version :
Processor : x64
Crash Address : ntoskrnl.exe+3dda20
Stack Address 1 :
Stack Address 2 :
Stack Address 3 :
Computer Name :
Full Path : C:\WINDOWS\Minidump\080920-14812-01.dmp
Processors Count : 16
Major Version : 15
Minor Version : 19041
Dump File Size : 1.434.428
Dump File Time : 09.08.2020 13:14:38
==================================================
 

Svet

Well-known member
Administrator
Joined
Nov 20, 2003
Messages
75,922
Yes, RMA the card or if you can test it to be sure in another computer 1st.
 
You must log in or register to reply.
Top