Encountering sporadic startup problems with the 5700X3D?
Encountering sporadic startup problems with the 5700X3D?
I improved my CPU to a 5700X3D a few months back and have faced some strange, sporadic problems that have become bothersome. Initially smooth, the issue has appeared intermittently over the past couple of months. It often disappears completely for long stretches, then reappears. The main signs are: sometimes the system won’t start at all, other times it takes a long time to boot (minutes), with the power indicator lit but no disk activity and no display output. Restarting or clearing CMOS hasn’t helped much; success seems unpredictable, though it’s more likely after the machine sits idle. This pattern also occurs when waking from sleep, though I’m not sure if that’s connected.
Other observations: When shutting down, the PC behaves normally—no errors in Event Viewer except for a few GPU driver crashes. Performance remains strong, temperatures stay low (CPU under 50°C, GPU around 60°C), and everything functions perfectly when operational. I haven’t experienced any crashes, BSODs, or driver problems since the upgrade.
I’m seeking advice on this behavior. Any insights would be greatly appreciated.
Additional details: The CPU model is a 5700X3D, installed on an ITX/ax board with ASRock B550 Phantom Gaming BIOS (version 3.40). I used 2x16GB G.Skill FlareX 3200C14 memory at various speeds and voltages, including a custom 3800OC configuration. RAM was optimized via Curve Optimizer, but resetting everything back to defaults didn’t resolve the issue. The motherboard lacks diagnostic tools, making troubleshooting difficult. I’m considering whether memory controller problems or RAM incompatibility might be at play, especially given the long boot times and eventual stability post-shutdown.
I also want to know if any screenshots of ZenTimings or other config data would help. My system specs are as follows: 5700X3D ASRock B550 Phantom Gaming ITX/ax, BIOS 3.40, 3800C14 memory at 1.45V, 1900 MHz clock, SOC 1.1V, stable for about three years. If you have any further information, it would be very useful.
Have you explored the updated BIOS releases? What is your CMOS battery level (without using measurement tools)? Is the SSD running the newest firmware? The hard drive status is normal (Hard Disk Sentinel).
Great questions! I haven't tried any newer BIOS versions, partly because of the boot issues making updates difficult (getting into the BIOS at all is a real hassle), and partly because both newer versions are betas. Neither are good enough reasons to not try them though, so I'll do that. I didn't have time to install any new software yesterday, but the SSD has the newest firmware according to Adata's software, and was sitting at 94% (IIRC) health there + in CrystalDiskInfo. SMART data also shows no errors, critical warnings or similar. Also, the CMOS battery thing really stuck in my mind since reading this last night - mostly because I've noticed the PC's clock being way off after some of these boot failure cycles, which is really a kind of red flag now that I think about it. Funnily enough I've never had to replace a CMOS battery through my 20+ years of building PCs, but I guess there's a first time for everything. Unfortunately not only does this board use a wired plug-in battery instead of an easily replaceable one, but it's stuck in between the rear I/O, under a shroud and behind the VRM heatsink, making it incredibly difficult to access (I doubt I can even get my multimeter probes in anywhere useful without removing the board from the case). Good thing I built my water loop with QDCs, otherwise this would be a very major undertaking. Still, since the battery isn't even mentioned in the motherboard manual I'll have to get it apart to identify what kind of battery it is (hopefully it's non-rechargeable and I can just jerry-rig in a CR2032 replacement instead of having to order something). Still, given the symptoms this is definitely something worth looking into, even if it's more work than I'd like.
Avoid any overclocking or undervolting until the problem is fixed. Stick to default speeds only. This prevents unnecessary complications if unstable OCs are present. Battery care is important too—better safe than guessing. As with Ryzen setups, beta updates often address one issue at a time, making it worthwhile to test again.
It was quite a challenge, especially since the connector was deeply hidden beneath the heatsink and IO shield assembly. Once it was out of the way, things stayed unclear. Initially, the battery looked normal at 3.17V under standard DC testing, but dropped to 2.85V when checked with my multimeter in battery test mode—normally a small load is applied. Compared to a fresh CR2032, it read 3.27/3.18, showing a much steeper drop under load, though not enough to confirm total failure. I plan to swap it out anyway, knowing the old unit was permanently bonded with Kapton tape and heatshrink. This situation has been familiar since I began troubleshooting; it suggests either a RAM incompatibility or a controller issue. The interesting part is this RAM has remained stable at 3800C14 for years with the 5800X, yet here it doesn’t boot at JEDEC speeds. While BIOS problems are possible, the 5700X3D was only slightly downclocked from the 5800X, giving them ample time to fix such issues before I arrived. No major faults have been reported in the three years prior to this setup. That’s reassuring. I’ll test the new battery after a short period with the PC and try adjusting settings one at a time to avoid overcomplicating things. Hopefully, this will make it easier to access the BIOS this time.
Well, I experienced a bit of an unexpected issue when the PC wouldn't start after reassembly and wouldn't power down even when I pressed the power button. Thankfully, it turned out to be a simple oversight—I had left both the PCIe riser and EPS connectors unplugged during setup. This should happen if I keep up with my PC maintenance routines late at night. For now, the new battery helped the system boot in a reasonable time, though I’ll need to test further before deciding. I’m holding off on updating the BIOS for a few days to check stability, then I’ll proceed. If that doesn’t work, I’ll acquire a CPU today and grab a RAM kit tomorrow. I’m not sure if I’ll replace the motherboard, but I’ll be ready when I am. Edit: After several restarts and power cycles, it seems to function properly at least temporarily. I even accessed the BIOS directly via ASrock’s tool, which had previously failed.
This update confirms the battery swap resolved the problems. No more boot issues or strange behavior after sleep. Appreciate the assistance—next time I’ll include checking the CMOS battery in my troubleshooting steps.