Start a Conversation

Unsolved

P

10 Posts

4910

August 13th, 2018 17:00

5060, freezing with AMD R5 430

We have several Optiplex 5060 systems (20+) that are randomly freezing when they have a AMD video card installed in them.   When the card is removed the systems stop freezing.   We have tried different AMD cards with no changes.   We have updated drivers and BIOS with no luck.     Our 5050 systems are having no issues...  Has anyone else experienced this issue?

367 Posts

August 13th, 2018 17:00

What AMD GPU is it? You do know that old ones don't support UEFI and will not display anything until the boot process is done (you're in windows).

Also if you are using an AMD GPU that needs auxiliary power and you don't plug in the PCI-E power cables the PC will work briefly but then stop due to a lack of power on the GPU.

Those could be what is happening but providing the GPU model could pinpoint the issue better 

10 Posts

August 31st, 2018 13:00

The AMD card is a R5 430.  We just purchased these cards with the computers.  They are plugged into the Blue PCI-E slot.

367 Posts

August 31st, 2018 16:00

I think I know what is happening.

Personally, I have used an AMD Vega 56 on a much older Optiplex 780.

I had a similar issue with the PC not working with the GPU I found that disabling the integrated graphics via the device manager (using the integrated graphics) and disabling the primary video auto option in the BIOS fixed the issue. Also with Radeon GPU's for some reason, it makes virtual monitors so if the PC boots and then the screen goes black press Windows Key+P on the keyboard a few times then the monitor should display the login screen.

 

10 Posts

October 22nd, 2018 10:00

Just replying that this is NOT the issue.  I have tried specifying one and then the other with no luck.  The ONLY solution is to replace the card with an NVIDIA or remove the card completely. 

Dell has been able to reproduce the issue but says it is related to hardware acceleration in browers.

Matthew

10 Posts

October 22nd, 2018 10:00

I have been in contact with their support engineers.  They CLAIM it is an issue with hardware acceleration in browsers but we have users who experience freezing not in a browser.  Are you experiencing it in browsers or ANY application?

1 Message

October 22nd, 2018 10:00

I have a client that is experiencing the same issue - freezing up at random times.  They also have a 5060 with the AMD R5 430.  Were you able to resolve this?

January 28th, 2019 01:00

I know this is old but...

 

I actually had this happen to multiple Dell systems in multiple offices. All r5 430's. Long story short, tried dell driver, windows driver and multiple from AMD while using DDU every time to clean out the old drivers etc. Dell support couldn't fix it and after billing clients for many hours trying to resolve, I had to just return the cards/eat it and buy NVIDIA 1030's which have been in production for months now and still working great.

To me it seemed like the r5 430's were pulling too much power. They would crash on remote sessions, any programs, stress tests, didn't matter. Comp would completely freeze and you would have to hard reboot it every time. 

 

TLDR; No one could fix the issue with any driver, had to return/eat cost and buy NVIDIA 1030's or something similar. 

1 Message

January 28th, 2019 09:00

We also have 5060s  freezing.  Tested  close to 100 of  them  so far and  have  seen  freezings  in  about 20%. 

Taking  out  the  AMD Radeon RX550 card  helps but doesn't seem to prevent the freezings. 

Dell  is looking into it and  we  are  awaiting their  analysis.

November 18th, 2020 02:00

I'm so glad I finally found this thread! I thought I was the only one having this problem!

My company has been plagued by this issue with AMD R5 430 cards on our fleet of around 500 MT and SFF Optiplex 7050s, for a few years now. I thought I was going crazy, and Dell only offered to try to remote control the PC when it was in the 'frozen' state, along with the ususal 'have you updated your antivirus software recently?' scripted solutions. Not very helpful.

For us, it's not a repeatable or predictable problem, which makes it hard to try to test settings/tweaks, but I think I've made some improvements over time.

Our PC's run in many different use cases. Some are normal office PC's with browsers running, others are fixed appliances which run full screen software, and others still do very little, and may sit at the logon screen for days/weeks at a time. It's not just browsers or GPU activity causing problems in my experience.

I suspect there are 2 stages of failures with these cards.

1. Animated 'Snow' artefacts on screen with a little tearing, but otherwise a working system. Snow is visible via VNC and other remote control utilities as well as on DVI outputs on the card. It almost looks like faulty RAM on the card, but it clears after a reboot, or the card cools down.

Optiplex 7050 with R5 430 showing snowOptiplex 7050 with R5 430 showing snow

2. Full system freeze, with a static blue or grey background. No processes are running. Anecdotally, checking the temperature of the air blowing out the back of the PC is around 10 degrees hotter than other PC's which haven't frozen. Occasionally you see remnants of what was shown on screen before (such as the Windows 10 clock on the logon screen).

Optiplex 7050 with R5 430 completely frozen, with solid grey screenOptiplex 7050 with R5 430 completely frozen, with solid grey screen

I'm not sure, but I think that the Snow effect is a precursor to the full system freeze, but I haven't witnessed that myself.

When I've seen a PC with the snow effects, I've run various utilities to check GPU fan speeds and temperatures while it's still operational. What I've seen is that the GPU fan seems to stay running at 30% at all times and doesn't ever automatically speed up/slow down, even under high load. I've managed to make the snow effect go away by manually turning the GPU fan speed up to 100% using AMD versions of the display drivers, rather than Dell versions, and watched the temperatures drop by 15 degrees in a few minutes. This might be a fix, but I haven't found a way to fix the GPU fan speed to 100% across an entire fleet of machines.

One other thing I've tried is to turn the CPU fan to 100% in the BIOS. It's loud, but that's not an issue for some of our PC's. That seems to help reduce the video cards temperature by a few degrees, and seems to stop it happening quite as often. It's a workaround, not a fix, but I've been able to script this via the Dell DCU utility, which I've rolled out to some machines.

The most recent thing I've tried is to try to edit the R5 430's firmware, or flash different versions of firmware for the card that I've found online in less than official sources. I was hoping that a different firmware version would have fixed the faulty (in my opinion) GPU fan speed settings. None of my attempts have been successful, and I've managed to brick a couple of spare cards attempting this. I don't recommend it, but I think the only true fix for this issue will involve updated card firmware.

The end result is that we're taking the R5 430 cards out of critical machines, and dealing with the other issues we have with the onboard intel graphics card (which are related to our use case more than anything else).

I'm sure Dell are still offering this card in their latest 7080s, so I'd suggest avoiding those. We've gone for some nVidia cards in our latest order, and some HP machines to see if the grass is any greener on the other side of the fence.

Hope this helps someone.

9 Legend

 • 

47K Posts

November 18th, 2020 04:00

"Animated 'Snow' artifacts on screen with a little tearing, but otherwise a working system."

This is clearly overheating GPU.

Once the damage is too bad it crashes.

I would be curious to see if the heat sink is bending the card in such a way that it fails.

There has been no recall of these video cards. Per the purchased Dell warranty period, you should contact your Dell TAM (Technical Account Manager). Provide to the TAM all of the effected system service tag numbers. The TAM should issue video card exchanges. Any Home users should contact the normal Technical Support channel to have the video card exchanged.

HEATSINK FailureHEATSINK Failure

 

 

 

February 5th, 2021 09:00

Did you get this fixed?  have the same issues now Feb 2021 with Optiplex 5080. Dell can't find the issue. 

 

Thanks, LFitz

February 5th, 2021 10:00

Nope, the main issue seems to be that the cooling doesn't kick in hard enough (or at all), or that there is a poor thermal connection between the heatsink and GPU itself.

To be honest, I've all but given up hope with these cards. They're not fit for purpose, and I've started removing them from machines that have them to reduce the number of support calls we get. We bought and paid Dell for them as extras with these machines, but they've got an almost 100% failure rate now, with every machine with one of the cards now having had these graphical issues.

Our workarounds did help slightly, so I'm positive it's a thermal issue of some kind. Even in 19 degree air conditioned data centres with excellent airflow, these cards still overheat while sat idle at the Windows 10 logon screen. It's just a bad design.

If I were you, I'd take the cards out of the machines you've got, and either return them to Dell, use the onboard graphics instead, or take the hit and buy new cards at your own expense. That's the best advice I can offer you based on my experiences.

Good luck!

No Events found!

Top