I have a problem with my Ryzen 7 3700X. My system is
Usually the system does not boot at all, i.e. it shuts down after
about 3 seconds and then immediately restarts, doing the same thing
until it gives up and remains off. After that, it does not react to
the power button, but immediately restarts (without pressing the
power button) after turning off and turning on the power via the
power supply button, doing the same reboot loop again. After not
touching the system for a day, it might boot into the OS, but then
randomly shuts down again while idling after only 1-2 minutes (it
remained on long enough for me to see the CPU temperature - it was
< 60°C and thus not overheating). I switched all components except
for CPU & Motherboard and thus hopefully excluded all error causes
except these two. I once "fixed" the problem by not touching the
system for > 1 week - after that, it ran for hours continually and
through multiple normal user-initiated reboots. But eventually it
shut down again while I was asleep, possibly due to overheating. I
suspect overheating, because I fully loaded 8 CPU threads (GPU was
idling), let them run for hours to test the system stability,
eventually saw the CPU temperature slowly creep to 90° (the
temperature in my room was probably rising), but went to sleep before
anything out of the ordinary happened.
I built/repaired over a dozen working PCs for myself and
professionally and the only thing out of the ordinary here was the
CPU temperature. I used the same bequiet cooler (have it twice) on an
i7-8700k (which has a 30 watt higher TDP) and never had any problems
with it - the Intel CPU idled at slightly above 30°C, never came
close to 90° even under full load, and changed temperature very
gradually. The Ryzen on the other hand idled at between 45°C and
60°C, fluctuated between these values in under 2 seconds constantly
and reached 80°C with just 1 thread loaded, still quickly
fluctuating about 10° C. That behaviour was consistent with both
coolers and both inside and outside a PC case. I cleaned the CPU
thoroughly with isopropyl alcohol before installing the new cooler. I
never overclocked anything (including RAM), I only changed fan curves
in UEFI. The motherboard unfortunately has no useful troubleshooting
tools (hex code display, piezo buzzer or something like that).
Any hints for the cause of the behaviour? I'm at a point where RMA
for the CPU or motherboard seems like the only option and I would
like to avoid that.
Thanks in advance.
Edit: Fixed formatting
- ASRock X570 Phantom Gaming 4
- 1x SK Hynix hma82gu6afr8n / 2x some Samsung ECC-RAM (probably M391A2K43BB1-CTD)
- MSI RTX 2060 Super Ventus OC / Gigabyte GT 710 (GV-N710D3-1GL)
- bequiet Pure Power 11 400w / bequiet Pure Power 11 500w
- Wraith Prism boxed cooler / bequiet Shadow Rock low profile
- nothing else is connected, except sometimes a USB key with my operating system (a Linux live system) and/or a USB keyboard
- CPU, Motherboard, Samsung RAM, larger power supply, MSI graphics card were brand-new, the remaining parts were used, but known good because they ran for months in my Intel system without problems
Usually the system does not boot at all, i.e. it shuts down after
about 3 seconds and then immediately restarts, doing the same thing
until it gives up and remains off. After that, it does not react to
the power button, but immediately restarts (without pressing the
power button) after turning off and turning on the power via the
power supply button, doing the same reboot loop again. After not
touching the system for a day, it might boot into the OS, but then
randomly shuts down again while idling after only 1-2 minutes (it
remained on long enough for me to see the CPU temperature - it was
< 60°C and thus not overheating). I switched all components except
for CPU & Motherboard and thus hopefully excluded all error causes
except these two. I once "fixed" the problem by not touching the
system for > 1 week - after that, it ran for hours continually and
through multiple normal user-initiated reboots. But eventually it
shut down again while I was asleep, possibly due to overheating. I
suspect overheating, because I fully loaded 8 CPU threads (GPU was
idling), let them run for hours to test the system stability,
eventually saw the CPU temperature slowly creep to 90° (the
temperature in my room was probably rising), but went to sleep before
anything out of the ordinary happened.
I built/repaired over a dozen working PCs for myself and
professionally and the only thing out of the ordinary here was the
CPU temperature. I used the same bequiet cooler (have it twice) on an
i7-8700k (which has a 30 watt higher TDP) and never had any problems
with it - the Intel CPU idled at slightly above 30°C, never came
close to 90° even under full load, and changed temperature very
gradually. The Ryzen on the other hand idled at between 45°C and
60°C, fluctuated between these values in under 2 seconds constantly
and reached 80°C with just 1 thread loaded, still quickly
fluctuating about 10° C. That behaviour was consistent with both
coolers and both inside and outside a PC case. I cleaned the CPU
thoroughly with isopropyl alcohol before installing the new cooler. I
never overclocked anything (including RAM), I only changed fan curves
in UEFI. The motherboard unfortunately has no useful troubleshooting
tools (hex code display, piezo buzzer or something like that).
Any hints for the cause of the behaviour? I'm at a point where RMA
for the CPU or motherboard seems like the only option and I would
like to avoid that.
Thanks in advance.
Edit: Fixed formatting
Last edited:
Facebook
Twitter
Instagram