@@A.S.Repairsi, I have Rx5500xt Nitro. The cards works, all the voltges are present and it' being detected in Windows(Error 43)/Linux (as secondary) but it does not give a picture. Now, I tried listing the devices in Linux, got the memory start address to run dmgg but it also says prefetcheable memory disabled. I've seen at lest two tutorials online, and there it doesn't say "disabled". Do you think it could be a BIOS issue (maybe mining card, I don't know it's history)? Or a memory issue (chips look brown and discolored, but mybe it's just a resulidue from thermal pads.. not sure. Thanks!
Did you come across a card that had all of it chips marked as bad? The same script is returning multiple results, all chips are detected as bad. BTW please put link to the script you used, there are multiple versions floating around.
I have msi rx570, screen goes black sometimes green when drivers are loaded, also artifacts. But in safe mode or without display driver it doesn't artifacts
HELLO BROTHER CAN U HELP.???? I have powercolor rx 580 4gb gpu the problem is the display had went dark while playing game and is not waking up from that state, the pc turns on but no fans spin no display just the fans get a little jerk while pressing the pc power button for turning the pc on. should i be doing this heat up method of vram . could this solve the problem ? i guess the problem is with vram i had it repaired from a local repair shop but the issue is still there after 20 to 30 mins the gpu again goes dark no fans spin no display but the pc stays on. also what could be permanent solution to this problem.
you are amazing brother keep up the good work, btw i had a question can we get a display in proper mining edition card rx 580? or are they only meant for mining >|?
Some of those cards have the place on the pcb for HDMI and display port which you can solder on yourself and change/remove the i/o shield. So yes, it is possible.
@@glorioussounds6865 In that case I don't think it is a mining card, I might suggest flashing the BIOS to the original file, you can find it in TechPowerUp
@@glorioussounds6865 you need to solder capacitors and resistors before the hdmi port because its not conected to the cpu and than it will work On the back side you will see the hdmi port contacts and right next to it the soldering pads for capacitors and ressistors
Hello please help me. I have run this script on my card i thought it shows correct but when i replaced the memories it still give same errors so i suspected that it is not working correctly when i put another card with no memories it was still giving same errors like with memories. So i am getting same errors all the time.
I am getting very small memory errors on chip B0 on my GPU. I am still using the GPU however as the memory fault is tiny and I don't notice it. Do you think just reheating this chip would fix this error or do you think I would need to replace it? I've seen videos where people are using a heat gun to solve issues with their GPUs. Willing to try if this might help.
Any problem with GPUs that heating would solve, cooling would also solve not to mention it's safer. If your GPU isn't crashing anywhere then I'd say just keep it as is or replace it if you really want to
@@A.S.Repairs hello, I have the same problem with a nitro rx 580 4gb, most games run fine, but in some games the lens flare is exagerrated and it results in circle-shaped artifacts (not vertical lines). I tried running occt vram test and I got over 7k errors, I think one of the memory banks has gone bad.
Nice Video Great Job I dont have PLL VCore Vrm voltages and not 1.8 voltage. I think my problem its APW 8722 its a protector comparator state for Pwm work and if gpu get any mal function this APW 8722 gone a helll.
I suffer from not running games and a problem with the graphics card, and I run it on another device and get an error message Error 0x887A0006 Unreal engine is exiting due to d3d device being hung and a problem with the graphics card itself. Is there a solution for Evga 1060 6g Ssc 6267 And I removed ddu the entire system and downloaded the latest version of NVIDIA and installed the latest version of DX12, and the problem still exists This error is in all new generation games and old generation games
Wow amazing! Never knew about the Linux trick to find the exact memory issue. I have a GTX 1660 with what I believe to have a memory issue as it has artefacts. I got this GPU in a system from Amazon and it come DOA i checked the bios and it had a different flashed bios than what model it is strangely. It still artefact on the screen unfortunately so I don’t know what else try next? Any help is appreciated
For Nvidia cards, you need to use another memory test tool called mats. I will make a video about it in the future but there is a guide on repair.wiki for it already.
I have rx 570 that sometimes throws black screen on driver install and when booting windows. But most of time it boots ok, games run but after a 10 min the colors start to get messed up and a little bit later the gpu throws black screen. Also the core frequency is lower than it should be. Bios is original, tried to reflash in case of corruption but its still same. Maybe its memory or its power controller or something? Is it even worth tryint to fix?
What is the maximum GPU temperature if you try to run Furmark for one minute or so (go for the 720p benchmark button). You may have bad thermal paste (dried up) or if the thermal paste is fine then you may have thermal pads a little bit too thick. This would be my first guess. Maybe the last time someone changed the thermal paste and/or thermal pads, it is possible to have one or more thermal pads missing (or it is very much possible to have them dried up). I'd say the VRMs can heat up a little bit too much if the pads are dried up and either blow up or the card is slowing down a bit. My other guess is that your card may have one or more solder balls ruptured and every time the card heats, it is happening to show weird colors, black screen or blue screen. It also can be a solder pad broken under one of the memory chips or even under the GPU itself and if this is the situation, a repair is not easy at all (this mean it can be expensive and it has to be a real professional who is doing the repair). Video cards usually have a tiny chip managing the video card's internal power supply. Components around that chip can be broken or even the chip itself can get broken or partially broken. There's not an easy answer to your question. The temperature of the GPU in the one minute 720p Furmark's benchmark should not exceed 80 Celsius. 75 Celsius should be the maximum. Sometimes the cooling fluids are gone from the heat pipes of the cooling system. I've seen this more often for the MSI RX cards. Like I said, there is no easy answer.
Hello brother, I am repairing the graphics card. Is it possible to convert an 8gb card to 6gb in the rx series? For example, the chip is faulty, even if the d0 and 01 vrams are changed, it still shows faulty. In this case, is it possible to disable d0 and d1 and run the card as 6GB?
thank you very much for shared this information, is very useful. I have a gtx 1060 strix oc wiht all voltages presents but no show image. I already checked all pci-e lines and are good, i cecked pex rst and is fine too, so, Why is it not showing video? I should mention that I bought this card with a short on vmem, I found it on a capacitor and removed it. the resistance in vmem is 65 ohms... pd: sorry for my bad english
I have rx 570 and stuttering in few games like a lot of stuttering just unplayable games like ac origins but odyssey running okay warzone 2 but warzone 1 running okay counter strike go 80 fps with stutter and this happening with every settings
Thanks for the guide video. I tried it with 6800xt. 10mb works fine with no errors. From 13mb onwards, there is noise on the screen and an error appears. Is any problem?
wow very useful I have rx 560 4GB which have code43 crash on load when sometimes driver gets loaded and also vertical lines sometimes in bios when booting is it the core or the memory
I have a rx 560 4gb which gives vertical lines artifacts only once in 10 boots and and when it does not I works 100% fine for indefinitely can you help me what might be the problem I have already tried your script and it was unable to tell what bank was faulty
Hi i have followed your method on my Gigabyte RX 580 4 GB Model. And the results i get are Number of Faulty Chips: 4 chip 3 and/or 4 is faulty at address : 0 chip 1 and/or 2 is faulty at address : 512 chip 5 and/or 6 is faulty at address: 1024 chip 7 and/or 8 is faulty at address : 1536 Does this mean all the chips are bad ? The card gives black screen and 100% fan spin as soon as i try too Start Furmark or any type of load.It crashes out.
Hi, what if all memory banks reads all error or failing, i dont think its the vram chips itself, the error reads same digits numbers, where could it be? The controller or the bios??
@@A.S.Repairs earlier i tried to replace the bios chip with different one, and the vmem is stabled in 1.5v now, and keyboard is now active, unlike previous it doesnt active the keyboard, took 30 secs to light the keyboard, now the problem is i cant flash it cuz i used a different model of bios chip came from rx 550,😅 do you think its the bios chip problem?
hello i have the sapphire rx 580 4gb which has vcore a dnd memory but no 1.8v no memory controller or pex any ideas or help would be appreciated excellent videos by the way very helpfull.
Retested using intel GPU too boot linux, No errors detected on RX580 Memory at e0000000 (64-bit, prefetchable) [size=256M] Memory at f0000000 (64-bit, prefetchable) [size=2M] number of faulty chips= 0 Total bytes tested: 4*67108864 Total errors count: 0 - every 268435456.0 OK: 268435456
@@A.S.Repairs still crashes if i set core and mem sliders to minimum. Default clocks GPU 1355 MEM 2000. (have dual bios and other is set to GPU 1380 MEM 2000)
@@A.S.Repairs it not good to do it that way. I do it before i buy board holder from bga station and preheater. With it u can desolder all mosfets or rams without interupting. Way easier work. All from ali 50e holder, 100e preheater
