End screen displaying a dark background with the word "Death" in bold text.
End screen displaying a dark background with the word "Death" in bold text.
So here's what's up, A little bit over a year ago I decided to build my own PC and to make it short I did. I watched a lot of yt videos on how to build it, followed every step by step and even took an inspiration from a build I saw online and followed their steps, everything was brand new except the graphics card which was my fathers colleague's which as he states was barely used and even had a warranty (still does). Upon installing windows and all the drivers and software's needed everything looked fine until I got that first "black screen of death" Im not sure whet it first happened or how but I thought it was nothing more than just some freeze or a bug. I clean installed the drivers it somewhat went away but as time passed the issue returned. At first I noticed how sometimes while watching youtube the video's would start to stutter (sound as well) for a duration of about a 1.5 sec. and by the time I could react or see what is happening it would be over, and sometimes I could see weird graphical glitches such as white boxes for no longer than 0.5 sec on the web and other apps such as Office Word. It got really bad at some point that it was almost impossible to play game's and after that I started poking around. I went into bios (even though I already beforehand enabled xmp etc.) and completely reset it to default did all the drivers again with a clean install and I even installed a bunch of software such as HWmonitor and other to check my gpu bios. No issue was found everything worked fine (on paper), no overheating etc. until I downloaded a free benchmark test (I'm not sure but I think it was UserBenchmark). And as a normal person I started doing all the tests including cpu and gpu and found an interesting problem. The test ran from dx9 all the way to dx11 and as soon as it got to the dx11 (space jellyfish test) 9 out of 10 times at the exact same point it went to the black screen of death with max fan spin and display not found and after every time i had to restart the PC since there was no way out of it (I tried unplugging the DP and plugging it back in etc.), nothing worked except restart. So to get to the point of it someone please help me as i really don't want to spent another 1000+$ on a new gpu im not sure if its software related or is the gpu just bad. I'm gonna list the pc specs below. -Also I have a 850W gold plus Psu and nothing was overclocked and the cables going to the gpu are two separate cables (as it should be 2x8pin connector) and even though I considered the psu to be a possible issue the benchmark test persuaded me otherwise. -MBO q-code shows no errors except one time the pc didn't properly boot which was a code 40 from returning from sleep (I now have disabled the sleep option and the issue is still there). Here's the full list of my pc specs: MBO- ROG STRIX X570-E GAMING WIFI II CPU- AMD Ryzen 9 5900X RAM- 2x G.SKILL 16GB Trident Z Neo DDR4 3600MHz CL16 GPU- TUF Gaming GeForce RTX™ 3080 V2 OC Edition 10GB GDDR6X (supported by a bracket so no gpu sag ) PSU- Cooler Master V850 SFX GOLD (SFX because the case is Lian li O11 dynamic mini) STORAGE- Samsung 980 Pro 1TB Cooling- 9x Lian li sl120 Uni fans (connected to a Lian li Hub) CPU Cooling - Nzxt Kraken Z53 240 mm (The Lian li fans are mounted on the cpu cooler and the hub is connected to a cpu pin connector which via PWM controls the spin rate) DISCLAIMER: The GPU does have a warranty, However It was bought in Germany (by the colleague whilst he was staying there) and was afterwards brought to Croatia where I live so there's no possible way of me going to the retail store and getting a refund otherwise I would have already done so. The files attached are my DxDiag and HwMonitor (the temp are a bit higher than usual as the pc need a spring dust clean). Anyways Please if you can somehow help me or you had a similar issue and know what to do please reach out Thank you in advance. DxDiag.txt HWMonitor.txt .dmp.rar
Start by checking your Windows Event Viewer to confirm what happened during the black screen. This tool often reveals that the Nvidia driver failed and didn’t recover. Note the exact time whenever it crashes. Also, observe whether the PC is fully off or just inactive with a display still on. Pressing the Num-Lock key can help determine if the system is responding. On a working machine, search for 'Event Viewer' in Windows Settings and look at the logs under System. Look for entries indicating nvlddmkm stopped functioning. This suggests a GPU problem. It’s wise to verify your motherboard BIOS is up to date. Having spare components and another machine handy can speed things up. If you have another PC, try connecting the 4080 graphics card to see if it continues to fail—this usually points to a hardware issue. Artificing the GPU often means faulty VRAM, but a bad power supply can also cause similar issues. My last experience involved a 3080ti; after thorough checks, I found a damaged PCIe cable and had to return it for repair.
I've checked the event viewer before and now I haven't noticed any issues before the crash. It only displays an error: The previous system shutdown at 7:49:01 PM on 4/3/2024 was unexpected, or critical – the system restarted without a clean shutdown. This might happen if the system stopped working, crashed, or lost power suddenly. That's roughly what I see. If it happens again, I'll investigate further but it seems it isn't being recorded as an error. Before the black screen, I see a message saying: The start type of the Background Intelligent Transfer Service service was changed from auto start to demand start (47 minutes before crash). This could mean the PC is still running in the background, which is why I can still hear in-game music or my friends on Discord, though not always. Sometimes everything just stops abruptly. Regarding artifacts, those white boxes appear occasionally while browsing websites or YouTube, but they disappear quickly if you don't notice. I might try checking for loose cables or looking for damage on the cables to be sure about the PSU.
Hey, I think this might be a GPU with a dead or failing VRAM. The main reason I wouldn’t be sure is the absence of crash driver logs in the event viewer. Your description matches that pattern. Can you try installing the card on another machine to see if the issue appears? If it still works there, you’re likely dealing with a faulty card and can proceed confidently with a return.
You might also consider using MSI Afterburner and lowering the VRAM speed (possibly the core clock too). It won't resolve the issue, but it could help if underclocking the GPUs fixes it despite lower performance. This would just add another point to the RMA process.
Hi, sorry to hear that. I don’t have another PC available to install the GPU. I’m not very familiar with clocking or similar details, but I’ll reach out to the Retail store via email to describe the problem and see if they can help with a replacement card or alternative solution. If that doesn’t work, I’ll stick with this card as long as it functions, while saving up for a new RTX 50 series. Thanks again for sharing this information—it really aids in explaining what’s happening to the store.