[SOLVED] What is wrong with my Graphics Card?

[Ebiye Idumange Elton]

Hi, I got my GPU - a Sapphire Tri X OC R9 290 about a month ago fairly used at a great deal. It when I got it, I ran Furmark stress test and noted it's performance. It was okay and it reached a peak temp of 76c. I played games on it for a couple of weeks, with no issues. Two weeks ago, I decided to overclock the card. I followed this guide https://www.anandtech.com/show/7601/sapphire-radeon-r9-290-review-our-first-custom-cooled-290/5y and aimed for those settings using MSI Afterburner. My first attempt was successful, but other subsequent attempts weren't and it was very unstable. So I aimed lower and finally settled for an overclock settings of 1060MHz (base clock was 1000MHz) and a memory speed of 1400MHz (base speed was 1300). I uninstalled Afterburner and used AMD Wattman to perform the overclock. I didn't touch the voltage settings since Wattman doesn't have that option. But I adjusted the Powertune to +10.
About a week ago, I couldn't play any games on it, due to the frequent crashes. And I was unable to properly shutdown my computer or set it to sleep. Each time I tried I would get the (this memory could not be referenced at Radeon.exe) I shutdown the computer using the power button.
I decided to reinstall Windows, thinking these problems were due to a driver update or Windows Update. With a newly installed Windows and a previous driver, I started seeing artifacts on the screen. And games would crash almost immediately - this was in stock settings. I hadn't applied any overclock to it yet. The artifacts appeared every time Windows was booted up.
Expecting the card to be dead, as a last resort, I decided to open up the card and inspect the thermals. The thermal paste applied to it was still moist, and there was no signs of viable damage. I applied a little thermal paste to the card and installed it. It didn't crash this time, with no artifacts appearing and I was able to game for a while. But it ran a lot hotter than before. Rising from a peak of 81c before to a steady range between 88 to 95c. I got some thermal paste and applied to it, some days after and yet the problems persisted. I've inspected the thermal pads and it all seems to be in order, although a bit old.
I've updated the BIOS of the card. And it still runs hot. Running Furmark now pushes the card to a peak of 96c, while earlier, the max Furmark could push it to the was 81c.
I've noticed that a pin in the GPU die is missing.
I'm open to any and all suggestions. I'll update the thread with more details and pictures.

The P.C specs are;
Intel core i7 4770s
16gb RAM
600 watts 80+ Bronze PSU
Barracuda 3TB hard drive
Barracuda 1.5TB hard drive
ASUS B85M motherboard

This is a link to the motherboard problem I'm having currently with the GPU.

[SOLVED] - Desktop not performing POST, or booting to BIOS, displays black screen with white underscorei

Hi everyone, my desktop doesn't boot to BIOS, or starts POST. I was having an issue with my Graphics card - it's a Sapphire Tri X OC R9 290. So in order to test the problem, I'd take the GPU in and out of the system and turn it on, in order to test if it was working. This morning after...

forums.tomshardware.com

 

[hang-the-9]

Unless you have a low quality power supply, I would say the video card has issues. Used card, you overclocked it which causes more stress on components, now it's crashing. Not a big stretch to see where issue may be. Unless you want to replace the stock cooler with something better to see if that helps, but at that point you may as well just replace the card with a new RX 580 which should be same speed.

 

[Dreamevil55]

Try the card in a separate system like on a friend's one and see what happens ?

 

[ElectrO_90]

Several things spring to my mind.
you killed the card with an overclock - because it is an old card (6 years old)
Fans are not working efficiently like before, and with you moving it about, may have upset it.
You haven't tightened the screws down on the heatsink properly so not enough pressure to let the heat escape
When the fans are on, are the spinning at 100% (do this in afterburner) do the fans have any strength in them? try and slow the fan down with your finger. if no power, then fan is dying.
Get a hairdryer without the heat option and point it on the gpu and see if that drops the cooling down because of the extra airflow (yes it will be noisy but just a test).

 

[Ebiye Idumange Elton]

Several things spring to my mind.
you killed the card with an overclock - because it is an old card (6 years old)

Well, I thought that cards could last till they become obsolete. The overclock I applied to the card was already tested and vetted to work on those type of cards by the team at Anandtech, so I didn't think it would be very dangerous. When I noticed that it was unstable at those settings I reduced the settings and such.
And I thought overclocking a card was okay.. as long as it done appropriately with the right settings and such.

 

[Ebiye Idumange Elton]

Fans are not working efficiently like before, and with you moving it about, may have upset it.
You haven't tightened the screws down on the heatsink properly so not enough pressure to let the heat escape
When the fans are on, are the spinning at 100% (do this in afterburner) do the fans have any strength in them? try and slow the fan down with your finger. if no power, then fan is dying.
Get a hairdryer without the heat option and point it on the gpu and see if that drops the cooling down because of the extra airflow (yes it will be noisy but just a test).

I'll verify the efficiency of the fans. But it seems okay. The card maintained normal temps before, so if the fans went bad, it must have been recent. The fans do seem to be okay. At a hundred percent, it's quite noisy.

I have tightened the screws on the heatsink as tight as it could be.

I look into procuring a hairdryer and testing it with it.

Thanks for your suggestions. I'm open to more.

 

[Ebiye Idumange Elton]

Try the card in a separate system like on a friend's one and see what happens ?

Thanks a lot for the suggestion. I'll try to do that when I can. Let me know if anything else comes to mind.

 

[Ebiye Idumange Elton]

Unless you have a low quality power supply, I would say the video card has issues. Used card, you overclocked it which causes more stress on components, now it's crashing. Not a big stretch to see where issue may be. Unless you want to replace the stock cooler with something better to see if that helps, but at that point you may as well just replace the card with a new RX 580 which should be same speed.

Thanks for the suggestion, I really appreciate it. The PSU I have was one included in a prebuilt HPE H9 1010a. I don't know the name but it's a 80+ Bronze PSU. You know, I'm quite concerned about how much stress a 6% overclock could cause to a card. When the card was okay. With the 6% overclock the max temp while on the Furmark stress tool was 84c, and while gaming was 81c. So at the time, it was well within safe operating temps. It was later on that the card started behaving differently.
I'm sorry, I just don't understand the fickleness of electronics.

Thanks for your suggestion, please I'm open to more.

 

[Ebiye Idumange Elton]

Unless you have a low quality power supply, I would say the video card has issues. Used card, you overclocked it which causes more stress on components, now it's crashing. Not a big stretch to see where issue may be. Unless you want to replace the stock cooler with something better to see if that helps, but at that point you may as well just replace the card with a new RX 580 which should be same speed.

You know, I was wondering, cause it appears that the motherboard of the P.C I was using just died. Either that or the PSU I'm using is faulty which is preventing it from powering on the equipments. I'll update the reply with the link of the thread I created on the issue. What I'm wondering is; if it is the PSU that is the problem, would getting a better PSU resolve the issue I have with the graphics card?
I'd be happy if that turned out to be the case. But please let me know what you honestly think.

 

[hang-the-9]

Too much heat is not the only reason a video card can fail, simple age can cause issues from expanding/contracting components, running the card faster than stock.

Only way to see if the card or something else is the issue is to test the card in another system with a known good power supply and see if it works there for a while.

 

[Ebiye Idumange Elton]

This thread is pretty old but I'd like to state that when the card was put in another system with better cooling, the R9 290 didn't heat up so much. Although it died about six months later. So it seemed that the issue I was experiencing was in part because the card was failing and because my P.C case didn't provided adequate cooling for that card.