RX580 GPU issues are limited to specific systems.
RX580 GPU issues are limited to specific systems.
I've owned an RX580 with 4GB RAM for several years, never used it for mining, and purchased it brand new. I ran it on an ASUS Z77 board with a 3770K that was overclocked to 4.2 briefly. Everything functioned properly. Recently, I received a donated HP Omen PC equipped with an HP board and a 4770K. I removed the internal components and rebuilt it, upgrading the GPU to 4.1 or 4.2, and installed the RX580. It comes with a 700-watt power supply from a Dell server. I don’t recall the exact time I started experiencing problems, but I first noticed my monitor (1440P ultrawide on DisplayPort) would flicker and switch settings unpredictably—changing color modes and occasionally flashing purple when watching YouTube. This happened sporadically, sometimes lasting days without issues. Gaming performed well overall, except for occasional crashes that usually meant a power cut or restart, occurring roughly every 4 to 7 days. Sometimes after a crash, the system would refuse to restart and shut down during Windows boot after logging in. Eventually, I decided to investigate further. During stress tests, I observed the GPU temperature rising into the high 80s. I refreshed the thermal paste, cleaned the heatsink, and managed to bring it down to 79–80°C under load. In subsequent tests, the PC would frequently fail by powering off abruptly or restarting. At one point, after a restart, temperatures reached 81°C. There were no blue screens or prior instability—just sudden shutdowns. Today, when I used the PC, MSI Afterburner monitoring stopped and displayed a distorted image on the far right of the screen. I concluded it was finally over. I returned the RX580 to its original ASUS Z77/3770K build—it worked perfectly, with no flickering, stable stress testing, and no issues. I also tried swapping in an R9 380 card into the HP Omen 4770K setup; it performed flawlessly. No flickering on the ultrawide, stable stress results, or any instability. If the problem was a power supply, I’d expect issues with the R9 380, but not with the displayport cable or the motherboard/PCIe slot. At this stage, I’m focused on diagnosing one or both machines to decide whether to sell or donate them. I’m ready to move on to a new system and ensure I resolve this before handing something back.
I experienced a comparable problem on an MSI RX 580. It seemed to be caused by temperature issues. Occasionally it would crash or freeze, particularly during gaming. The machine was housed in a solid front case with a cooling fan and reached very high temperatures quickly. I’m not sure how long this continued before I noticed the signs. It’s possible heat damage occurred.
These signs could align with both VRM and memory overheating issues. It’s possible that excessive heat affecting VRMs might lead to instability or shutdown, while memory-related overheating could also trigger similar symptoms. It’s worth checking if the issue persists across different components.