Jump to content

Posting on behalf of a friend.

Just built a new system 

 

7800x3d

Msi 4090

850w corsair psu

Asus b650-a

 

But it crashes under sustained gpu loads. The crash has been successfully reproduced using heaven benchmark and typically happens between the 8 and 15 minute markers.

 

* Using hwinfo I can see no indication of high temperatures cpu is below 80 c under stress with cpu-z and gpu below 70 c while in heaven benchmark. 

*Inserting an old gpu (1070) yielded no crash after 20 minutes.

*All psu connections have been removed and reconnected. 

*Fast boot disabled

*ram removed and shifted to alternate a2 and b2 slots

 

 

Calling out to any ideas or recommendations on what to test next or any possible solutions to this strange issue.

Link to comment
https://linustechtips.com/topic/1552889-new-build-crashes-under-sustained-gpu-load/
Share on other sites

Link to post
Share on other sites

6 hours ago, Tr1ggerfish1 said:

Posting on behalf of a friend.

Just built a new system 

 

7800x3d

Msi 4090

850w corsair psu

Asus b650-a

 

But it crashes under sustained gpu loads. The crash has been successfully reproduced using heaven benchmark and typically happens between the 8 and 15 minute markers.

 

* Using hwinfo I can see no indication of high temperatures cpu is below 80 c under stress with cpu-z and gpu below 70 c while in heaven benchmark. 

*Inserting an old gpu (1070) yielded no crash after 20 minutes.

*All psu connections have been removed and reconnected. 

*Fast boot disabled

*ram removed and shifted to alternate a2 and b2 slots

 

 

Calling out to any ideas or recommendations on what to test next or any possible solutions to this strange issue.

I would try a better power supply. A 4090 can draw close to 400W under load, 450W if you push it. Then there are transients that can spike as high as twice the GPU's rated power for a few dozen microseconds, which would be enough to trigger the OCP of an underspecced power supply. How exactly does it crahs? Is it just an exit to desktop from the application running on the GPU or does the entire PC turn off?

Link to post
Share on other sites

35 minutes ago, DreamCat04 said:

I would try a better power supply. A 4090 can draw close to 400W under load, 450W if you push it. Then there are transients that can spike as high as twice the GPU's rated power for a few dozen microseconds, which would be enough to trigger the OCP of an underspecced power supply. How exactly does it crahs? Is it just an exit to desktop from the application running on the GPU or does the entire PC turn off?

The video output will freeze for arround a minute then it will bluescreen with code: DPC Watchdog Violation.

Link to post
Share on other sites

Update: the system started to cycle while booting into windows, windows was reinstalled with no issues however as soon as the system grabbed the video driver the system cycled with a black screen. On for roughly 1 second then black screen for roughly 5 seconds. system unresponsive but did not reboot.

 

We removed the power supply and gpu and tryed the gpu in another system (windows 10) same behavior on the other pc, once graphics drivers installed video output would cycle with black screen and system is unresponsive.

Link to post
Share on other sites

4090's are among the most powerhungry cards... i wouldn't install it with less than 1200w psu in my system... most would say 1000w.. but i've seen a few of those with issues too. 

 

i do run a 1600w psu now tho.. 😄 the less you stress a PSU the longer it will last and degrade over time. 

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×