Display black screen issue when GPU is being loaded
Display black screen issue when GPU is being loaded
are still active and the fans are spinning when it happens (including GPU fans). It only happens during heavier load. I can play certain lighter games (Trackmania for example) fine. Specs: 10700K ASUS 3070 Strix OC 16GB Corsair DDR4 running at 3200MHz 1000w ASUS PSU (overkill, but I bought this to try and resolve the issue) Doesnt seem to be overheating, as the CPU and GPU temps seem fine. No overclocking (except for factory OC on GPU). I've reinstalled drivers (using DDU), clean install of both W10 and W11, tried new RAM, even tried with a new motherboard. Latest BIOS version. I'm running two seperate PCIe 8-pin cables from the PSU into the GPU. I even tried to RMA the GPU but the store where I bought from it ran tests and couldn't reproduce the error, so they wouldn't take it back. In the store they ran Furmark for over 48h straight without any issues or temperature anomalies, but when I try to run Furmark on my own system it crashes. I've tried running Furmark with chassi fans at 100% for maximum airflow. Windows does not generate any log files, presumably because I have to shut the PC down by holding down the power button. Everything software related is up to date (Windows version, GPU drivers etc). I've also tried installing some older GPU drivers. I've attatched two GPU-Z logs from when I was running Furmark and it crashed. I've only included the data leading up to the crash, it ran furmark for about 5-10minutes before crashing. I'm leaning twoards a PSU problem, but I've already replaced it. It's 1-ish year old. Everything is as free from dust as I could get it. Any help or ideas would be appreciated. GPU-Z Sensor Log2.txt GPU-Z Sensor Log.txt
Hi, so I've been having this problem for a while now. My GPU(?) crashes during intense load. No BSOD, just a black screen and then I have to force a shutdown, the PC does not reboot by itself. All the LED are still active and the fans are spinning when it happens (including GPU fans). It only happens during heavier load. I can play certain lighter games (Trackmania for example) fine. Specs: 10700K ASUS 3070 Strix OC 16GB Corsair DDR4 running at 3200MHz 1000w ASUS PSU (overkill, but I bought this to try and resolve the issue) Doesnt seem to be overheating, as the CPU and GPU temps seem fine. No overclocking (except for factory OC on GPU). I've reinstalled drivers (using DDU), clean install of both W10 and W11, tried new RAM, even tried with a new motherboard. Latest BIOS version. I'm running two seperate PCIe 8-pin cables from the PSU into the GPU. I even tried to RMA the GPU but the store where I bought from it ran tests and couldn't reproduce the error, so they wouldn't take it back. In the store they ran Furmark for over 48h straight without any issues or temperature anomalies, but when I try to run Furmark on my own system it crashes. I've tried running Furmark with chassi fans at 100% for maximum airflow. Windows does not generate any log files, presumably because I have to shut the PC down by holding down the power button. Everything software related is up to date (Windows version, GPU drivers etc). I've also tried installing some older GPU drivers. I've attatched two GPU-Z logs from when I was running Furmark and it crashed. I've only included the data leading up to the crash, it ran furmark for about 5-10minutes before crashing. I'm leaning twoards a PSU problem, but I've already replaced it. It's 1-ish year old. Everything is as free from dust as I could get it. Any help or ideas would be appreciated. GPU-Z Sensor Log2.txt GPU-Z Sensor Log.txt
Raw_Noodles
12-02-2020, 10:13 AM #4Have you tried monitoring the 12V rail while putting the GPU under load? Use HWinfo64 and scroll down the sensors readings to the GPU and it'll be listed there. Also, you used the cables provided with the ASUS ROG Strix 1000?? Most modular PSUs use different cable wiring pinouts between the cable and the PSU, so there's a chance at the component end power is sent to a pin that's not supposed to receive power/or is lacking power. Thinking maybe the PSU is tripping protections as Ampere has high transient loads. This is a high quality PSU and shouldn't have any problems, but it's always possible that a new PSU is defective.
duhazneubin
12-14-2020, 02:54 AM #5You're using the provided power supply cables, and while the voltage holds up under normal load, it drops to zero during a crash. This could be either the main issue or just a sign of something else. The HWinfo log is attached for reference. When selecting a PSU, keep in mind factors like efficiency, thermal management, and compatibility with your system specs.
Sookabooyat
12-16-2020, 05:36 PM #6This seems to indicate a potential issue with the power supply. I suggest considering the Corsair RMx series, known for their durability and ability to handle high transient demands. Review the PSU tier list in the provided link and select between Tier A or Tier B based on your requirements. They also offer a 1000W PSU at RM1000, which could be a viable alternative if needed.
AwakeningTroll
12-16-2020, 11:40 PM #7Hey there! I noticed something intriguing. I haven’t swapped out my PSU yet. While experimenting with settings and raising the GPU power limit in MSI Afterburner to 125%, everything seemed fine—card drawing about 300 watts, temps staying at a safe 77°C. The cool part is: when I kept the 125% limit active for around 15 minutes, the system crashed instantly. If I switched back to 100% within a minute after starting the test, nothing happened. But once it hit 100%, the crash returned. It’s odd that the crash happens when power drops, but I’m not sure if this points to a PSU problem. What do you think?
Mel_Kawaii
12-17-2020, 01:14 AM #8It seems unusual for a GPU to behave this way. The power supply might be the issue, especially if minor power fluctuations affect performance. Alternatively, if stability drops at full capacity, the problem could lie with the graphics card itself. Let's explore both possibilities and see where it leads.