Error messages appear unpredictably during startup, even when CPU settings remain unchanged.
Error messages appear unpredictably during startup, even when CPU settings remain unchanged.
I built my own PC in 2019 and have faced some problems over the years. Likely the same one you're experiencing. The hardware includes a Ryzen 5 3600, an RTX 2060 Super from Gigabyte X570, a Corsair Vengeance LPX with CMK16GX4M2D3000C16 cooler, and a MasterWatt 650W 80 Plus Bronze power supply. It's getting a bit outdated but still works well for me. The main issue seems to be random crashes and error messages, sometimes involving Network drivers, GPU, or even Firefox. There was a discussion in 2021 about similar problems, possibly due to faulty RAM. I tried many fixes—new OS versions, updated drivers, BIOS tweaks, different settings—but for a while everything stayed stable. Recently, I experienced crashes again, this time linked to Xinput drivers and a new controller. After researching, I found a solution that resolved it. Now I'm curious: what exactly is triggering these problems? I’m not sure if adjusting clock speed and voltage alone will solve it, but I hope understanding the root cause helps. I’d appreciate any advice from others who might have faced similar issues.
Visit C:\Windows\Minidump to see if any minidump files exist. If present, return to the Windows directory and transfer the entire Minidump folder to the Downloads folder. Compress the copied folder and include it in a post.
It seems memory is involved here. When Zen 2 CPUs malfunction, they tend to trigger memory errors rather than CPU faults, suggesting a hardware issue is more probable. I’ll share my past copy-paste about memory as well. Memory isn’t always RAM, but it’s usually the main concern. Windows moves low-priority RAM data into the page file and retrieves it when needed, making storage appear like memory. The memory controller resides inside the CPU; a failure there will simply present as memory issues. When storage problems affect about half of the dumps, storage or its drivers are likely the culprit—this isn’t present here, so it’s unlikely. If any components are overclocked or under-voltaged, remove them. This includes verifying that Precision Boost Overdrive (PBO) is turned off in the BIOS. Your XMP profile is set to 3000MT/s, but the CPU officially supports 3200MT/s, so it’s not a true overclock. To verify RAM health, run the machine normally with one stick at a time. If only one stick causes crashes, that stick is faulty. If it crashes with either stick, the CPU is likely the cause. Memory testers often miss defective RAM, especially DDR4 and newer types, so I’m not confident in their results.
You think memory is the problem, so turn off all overclocks and run tests one stick at a time. If memory is the issue, setting a stable clock and voltage might resolve it.
It manages the speed well since it's only 3000MT/s, though it's safe to turn it off. And definitely test with a single stick at a time. We don't know if the problem is the same as before.
I mentioned adjusting the CPU to 4GHz and 1.25v, which keeps my system running smoothly. In reality, I’ve already resolved the issue, and everything is functioning flawlessly now. I’m just trying to understand why these specific settings are necessary for stability.
I faced a comparable problem with another setup. It seemed the system wasn’t receiving sufficient power during idle, or more likely, the motherboard’s transient response wasn’t fast enough—rarely triggering BSODs. Switching to fixed clocks or increasing the minimum clock frequency from 800MHz to 1600MHz resolved the issue. Additionally, my CPU’s idle power consumption rose from 3W to 6W (just a minor note). TL;DR, my motherboard’s VRM was likely aging and couldn’t handle the power spike quickly enough, leading to random idle BSODs.