• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

Only easy driver can detect the GPU.

Zweineys

New Member
Joined
Nov 17, 2023
Messages
12 (0.02/day)
The previous owner of sapphire pulse rx 580 said that it couldn't be detected anymore after uninstalling the driver and using DDU. The fan is still moving and and previosly, the driver date was still recent like few weeks ago before DDU (Picture 5).

So I tried it and it showed blackscreen if plugging to the card directly hence I'm using the mobo's port.

And yep, I checked it on device manager and gpu-z only to show my igpu only. AMD driver installer can't read the card too. Out of curiosity, I installed easy driver, and behold, it was there in unplugged section.

So while waiting for easy driver to download the driver, just in case anyone has a better idea to try or as a plan B? I heard that easy driver tends to download random driver but it should be compatible as AMD is still suporting this card with driver updates. (Can I install an older driver instead of the latest one with easy driver?)

What I already done,
-DDU with restart
-DDU with shutdown for new card
-Refresh and showing hidden device in device manager -> only revealing microsoft basic display adapter. (This perhaps is the card)
-Taking out cmos battery
-Checking for crack or damage in the card's pin
-dxdiag and wmic path win32_videocontroller get caption only showing igpu

PC Spec
-Asus B150M-AD3 w/ G4400 (Pcie x16 is fine as previously I tried using gts 450)
-PSU MSI A650BN 650 Watt Bronze
 

Attachments

  • 1700933241995.png
    1700933241995.png
    69 KB · Views: 96
  • 1700933278842.png
    1700933278842.png
    295.1 KB · Views: 86
  • 1700935013480.png
    1700935013480.png
    238.7 KB · Views: 81
  • 1700935158983.png
    1700935158983.png
    102.1 KB · Views: 86
  • 1700935269176.png
    1700935269176.png
    296.3 KB · Views: 87
Last edited:
I tried to click that link, and I got error UI.

Can I download from here instead? https://www.driverscloud.com/en/ser...ware-adrenalin-2020-20112-win7-64bit-nov18exe.

Also do I need to change my windows to windows 7? Currently, I'm using windows 10.

So I downloaded from driverscloud but I got the same error with the driver that I had tried earlier. It seems even this driver version can't detect the card.

Previous Failed driver
-amd-software-adrenalin-edition-22.11.1-win10-win11-nov15

Same errror.

Error 173 - AMD Software install cannot proceed as no AMD graphic hardware has been detected in your system


Sorry a little bit unrelated but did my previous reply need a moderator approval's and is invisible to normal visitors because I'm attaching a picture?
 

Attachments

  • 1700938677080.png
    1700938677080.png
    122.5 KB · Views: 48
Error 173 - AMD Software install cannot proceed as no AMD graphic hardware has been detected in your system
OK, 1st things 1st.
The driver i provided works perfectly on win 10 x64. At least it does on my computer.
The fact you are receiving error 173 - means that you have a faulty video card that is not even detected by system.
Before anything else please remove the card and clean PCIex slot. If you can use your phone camera and take a closer look at the slot pins.
After that reinsert the card in the slot.
If the problem with error 173 persist then you should consider to take the card to a professional repair shop.
 
Easy driver download is finally done but I got error during `The thid party inf does not contain digital signature infromation`. 0.6 mb/s limit in that platform is really painful. I was thinking to disable digital signature from safe boot or setting until I saw your post.
OK, 1st things 1st.
The driver i provided works perfectly on win 10 x64. At least it does on my computer.
The fact you are receiving error 173 - means that you have a faulty video card that is not even detected by system.
Before anything else please remove the card and clean PCIex slot. If you can use your phone camera and take a closer look at the slot pins.
After that reinsert the card in the slot.
If the problem with error 173 persist then you should consider to take the card to a professional repair shop.
All seems fine except I saw a light scratch that is hard to see except if you set the light source correctly.

Cleaned and still same error 173

More info about the card,
-Fan always 100% so I think either the cable is loose, problem with the driver or using mod bios
-Can't turn on the PC except if the gpu is off and then turn it on after bios screen by plug in the 8 pin power or using second psu.
-The bracket seems to bend a bit, it might cause improper sitting which I kinda doubt. Should I try to buy riser to confirm?
 

Attachments

  • 1700983678895.png
    1700983678895.png
    118.3 KB · Views: 62
  • 1700985343666.png
    1700985343666.png
    280.9 KB · Views: 54
  • 1700985388742.png
    1700985388742.png
    200.8 KB · Views: 56
After disabling the signature and then run the display driver, the card can be read with gpu-z now. But when I want to test the card with 720p furmar, it always crashes. Userbechmark too during gpu test. Screen becomes black, but I can still hearing cpu and gpu fan.

I tried to install adrenalin driver 22 instead of 23, but the problem is still persistent.

Had I installed the driver? I can only see amd software. I opened adrenaline and the driver version is changing too. If in Nvidia, I can see driver software in the software list and Nvdia control pannel.
 

Attachments

  • 1701000226214.png
    1701000226214.png
    943.7 KB · Views: 74
Last edited:
-Fan always 100% so I think either the cable is loose, problem with the driver or using mod bios
If that happens, then we have a dead soldier here.
I received that behavior only on dead cards. And trust me i had a few in my hands.
But i did not said anything about the PCI connector from the card, i was talking about the slot which is located in the mainboard. Check it out to see the pins.
On the other hand i would suggest to take the card to a friend and try it in his/hers computer.
 
After disabling the signature and then run the display driver, the card can be read with gpu-z now. But when I want to test the card with 720p furmar, it always crashes. Userbechmark too during gpu test. Screen becomes black, but I can still hearing cpu and gpu fan.

I tried to install adrenalin driver 22 instead of 23, but the problem is still persistent.

Had I installed the driver? I can only see amd software. I opened adrenaline and the driver version is changing too. If in Nvidia, I can see driver software in the software list and Nvdia control pannel.
I checked the event viewer and encountered these errors. I read that `The description for Event ID 56 from source Application Popup cannot be found.` this one is because of the PSU so perhaps I should try undervolt it.

I just knew if adrenaline has a test gpu feature and tried it. 2 things seem amiss for me.
-The temp can reach 87% C for 15-30s test.
-Fan speed doesn't seem to be accurate and I can't change the speed. Fan controller software can't do that either.

So I will try undervolt and checking the thermal paste later.

Could anyone please check my event viewer log in case I missed something?
 

Attachments

It seems like you are going through a lot of effort and weird/unrecommended software hacks to avoid coming to the realization that you have a hardware problem.

As recommended, I would put the same card in a different computer and see if it continues to misbehave. If it does misbehave, card problem. If it doesn't misbehave, motherboard problem.

I think it is odd to blame the fan wire, power supply, driver, and BIOS. Like every separate thing is wrong and you just need to solve each problem individually? nah
 
It seems like you are going through a lot of effort and weird/unrecommended software hacks to avoid coming to the realization that you have a hardware problem.

As recommended, I would put the same card in a different computer and see if it continues to misbehave. If it does misbehave, card problem. If it doesn't misbehave, motherboard problem.

I think it is odd to blame the fan wire, power supply, driver, and BIOS. Like every separate thing is wrong and you just need to solve each problem individually? nah
Because it feels rewarding when you can resolve an error message and it is fun to kill time on the weekend.

Sure, will give an update later.

Yeah, I am also really doubtful if it was the PSU. I'm more certain it is due to overheating. (Edit, even with UV and UC, during test, the temp can still reach 85 from 30-35).

For the fan problem, I'm only listing the possible cause for this from other threads since if only there was a report log for fan :D. That doesn't mean that I'm saying all of them are the problem since I haven't checked yet.
-It might be because the BIOS is tampered with to set the fan to 100% for mining
-There are several posts about fan problems for polaris card from driver 22.6 upward
-For the fan wire, I mean loose power pin or improper plugging.
 
Last edited:
Ded card.

Also, be careful slotting it into PCs willy-nilly. It preventing power-on is often indicative of a dead short or Major hardware failure in the card.
Thankfully, there was no damage elsewhere but, my last GPU that had same failure symptoms (9500GT) had a VRM explode while trying to get signs of life from the card.
 
So after cleaning and repasting, the card can run benchmark now. I tried superposition 1080p High and got around 40ish FPS with max temp 84 C.

The only problem left is the fan, always around 2950 rpm starts from the time I click the power button. The rpm can't be changed and even without a driver, it always runs on high regardless of the % speed.

So I was thinking of trying to reflash the bios. Does the sub-system id need to be exactly the same? I'm having difficulty finding it.
 

Attachments

  • 1701184641365.png
    1701184641365.png
    620.6 KB · Views: 77
  • 1701184699183.png
    1701184699183.png
    126.9 KB · Views: 82
  • Ellesmere.rom
    Ellesmere.rom
    256 KB · Views: 48
Last edited:
So after cleaning and repasting, the card can run benchmark now. I tried superposition 1080p High and got around 40ish FPS with max temp 84 C.

The only problem left is the fan, always around 2950 rpm starts from the time I click the power button. The rpm can't be changed and even without a driver, it always runs on high regardless of the % speed.

So I was thinking of trying to reflash the bios. Does the sub-system id need to be exactly the same? I'm having difficulty finding it.
Flashing the GPU firmware is almost never the solution. The more probably cause is a faulty fan cable (or fan), there are plenty of examples out there of non-working PWM fan control due to a problematic connection.

GPU-Z correctly identifies the card as RX580 with Micron memory, so leave the firmware alone imho.
 
Bet it's got a mining bios on it.
I'd try a flash.
 
Hello.
Indeed, you have an ex miner there.
I checked the BIOS and what you have right now is an aggressive one.
So yeah you need to reflash it.

Here you have a BIOS that matches your card ID sub ID and memory code :

Flashing guide :
 
Hello.
Indeed, you have an ex miner there.
I checked the BIOS and what you have right now is an aggressive one.
So yeah you need to reflash it.

Here you have a BIOS that matches your card ID sub ID and memory code :

Flashing guide :
So I tried this bios and it was successful in one try but the fan is still on high rpm and when I'm changing the driver, the bios became unknown once.

Reflashing it with your bios again and it is no longer unknown but no changes for the fan.
 
Well, what i do is to search for a BIOS that is suitable for your card.
On the other hand i do not know the status of your hardware.
I mean an ex miner will be an ex miner.
The hardware was abused.
So there is a small chance to repair your card by opening it and changing the thermal pads from memory and repaste your GPU.
But, as i said that is a small chance that should never be ignored.
 
Well, what i do is to search for a BIOS that is suitable for your card.
On the other hand i do not know the status of your hardware.
I mean an ex miner will be an ex miner.
The hardware was abused.
So there is a small chance to repair your card by opening it and changing the thermal pads from memory and repaste your GPU.
But, as i said that is a small chance that should never be ignored.
Reading your message about hardware is giving me an idea to try changing the whole fan body.

So I used a single 4-pin fan and tried it. It was kinda strange that the fan was running at 3200 rpm from the start and I also couldn't change the rpm.

I'll try the thermal pads at a later date.

Edit,
Any idea why there are thermal pads on the back of the card? Most tutorials are only adding 1mm row and 0.5-1mm to each memory block.
 

Attachments

  • 1701318833316.png
    1701318833316.png
    509.7 KB · Views: 49
Last edited:
Hello
Pads are mounted from the factory ONLY where they were designed to be.
And that is between hot electronic pieces and a cooler.
So having strips of pads on the back of a card put there with hand, and with no cooler to take the heat they transfer is useless.
I don't know the history of your card and neither do you.
Who knows what happened with it.
 
Hello
Pads are mounted from the factory ONLY where they were designed to be.
And that is between hot electronic pieces and a cooler.
So having strips of pads on the back of a card put there with hand, and with no cooler to take the heat they transfer is useless.
I don't know the history of your card and neither do you.
Who knows what happened with it.
I see, then I will remove them. The thermal pads will arrive in a few days.

Is there another bios that I can try in meanwhile? I tried to undervolt it by 25 and it crashed lmao. With the mining one, I can undervolt it to 1070.
 
The BIOS file i provided earlier is the right one.
You don't need to try all of them.
On the other hand let's wait for pads to arrive.
 
The BIOS file i provided earlier is the right one.
You don't need to try all of them.
On the other hand let's wait for pads to arrive.
After changing the thermal pads, while there is a huge temp down during heavy load, sadly the fan still can't be adjusted.

Just want to make sure, this means success right for changing bios?
 

Attachments

  • 1701539766838.png
    1701539766838.png
    205.8 KB · Views: 52
Hello.
Yes mate, the picture you've posted means that the flashing procedure was ok.
I don't know what else to do, except maybe some electronic pieces that are damaged.
Can you take the card to an repair shop ?
 
Hello.
Yes mate, the picture you've posted means that the flashing procedure was ok.
I don't know what else to do, except maybe some electronic pieces that are damaged.
Can you take the card to an repair shop ?
Sorry for the late reply. Bit busy with life due to the end of the year coming soon.

Remember the odd placement of thermal pads? There is one more on the front site of the board and since that piece seems of higher quality than the one in the memory, I didn't touch it. Out of curiosity, I pressed that part and didn't feel anything hard. I removed that part since I can't find a similar situation in any online resource. Under it lays empty space similar like this picture but I forgot the exact spot. For reference, that black thermal pad is really thick. Like almost twice the thickness of shirt's button.

Hence I tried to bring it to repair shop but since the shop didn't have the sparepart, it was returned to me and the mystery remained unclear.

Anyway since the card is working fine, I gave it away to my cousin as christmast present.

Thanks everyone!
 

Attachments

  • 1703165917806.png
    1703165917806.png
    430.7 KB · Views: 40
We are glad that the card is working as it should.
Merry Christmas !
 
Back
Top