[SOLVED] Random shutdowns - thermal, power, or what??

trainut

Reputable
Oct 15, 2014
27
0
4,540
I have been using a 2011 Dell XPS 8300 for the past 8 years, with periodic upgrades such as the addition of an SSD, a couple GPU upgrades, etc. Recently, though, I was really feeling the slowness of that i5-2400 and I decided it was time for a platform upgrade. I bought a Ryzen 3600, ASRock X570M Pro4, Inland Performance 1TB PCIe 4.0 NVMe, and 16GB (2x8GB) Corsair Vengeance LPX DDR4-3600 CL18. My plan was to use a few components from my existing build for the time being, such as the Dell case (I did a bunch of research to make sure the front panel/USB pinouts were compatible), the Gigabyte GTX 1070 I had put in the Dell a few years back, the Cooler Master Elite V2 550 (a ~485W PSU according to JonnyGuru) I had installed for the first GPU upgrade several years ago, and the SATA SSD and HDD from the Dell as secondary drives.

The last of the parts got here today and I installed everything without any issues. It was running well for about 7 hours while I installed Windows and all my programs, ran a few benchmarks, etc. About an hour ago it started suddenly powering off and restarting. My first instinct was to check the CPU temps, and while idle does seem a bit high (45-55C) it doesn't get too far above 75C during intensive use. I doubt it's a PSU issue, as the Ryzen 3600 actually has a lower TDP than the old i5-2400 and that was the only major change in terms of power draw. I already installed the proper drivers, newest BIOS, etc. which was all working fine for several hours. Event Viewer lists what is happening as "Event ID 41 Kernel-Power" which from what I can tell may point more towards a power issue. Any ideas?

EDIT: After one more occurrence of this during which I was watching HWMonitor closely, I noticed that the CPU temp is spiking like crazy very briefly (from 75-77C under full load to like 89 in a split second and then back down to 77 immediately) which is alarming, but still, shouldn't I be seeing thermal throttling at TjMax as opposed to shutdown (and isnt TjMax 100C)?
 
Solution
Today I went out and bought a Corsair CX650M. No change, the system still powers off suddenly. I am going to keep the new PSU since it was on my to-do list to get a better unit anyway, but this is a bit troubling.

I have also noticed that it tends to happen after a few minutes of sustained write to the NVMe SSD (mainly while downloading games on Steam). It has never happened during gameplay or benchmarking, just while writing to the drive.
Interesting. Unfortunately I am not super knowledgable when it comes to NVMe drives, however, that still strikes me odd that this happens when writing to the drive. This makes me suspect that there is something wrong with the motherboard, but I am not sure.
Try limiting the CPU to 70% speed in the windows power plan and see if that helps at all. If the temps fall that may have been the issue, although that PSU might have been on its last legs. 485w is a bit low for a 3600 plus a 1070 under load. Also, did you correctly install the AMD provided heatsink for the CPU? Did you flip the tension lever so it makes good contact?
 
Try limiting the CPU to 70% speed in the windows power plan and see if that helps at all. If the temps fall that may have been the issue, although that PSU might have been on its last legs. 485w is a bit low for a 3600 plus a 1070 under load. Also, did you correctly install the AMD provided heatsink for the CPU? Did you flip the tension lever so it makes good contact?

So apparently a lot of Ryzen 3000 series chips are not feeding correct data to apps other than Ryzen Master according to some threads online. I downloaded Ryzen Master and there aren't actually any of those crazy spikes. The real temperatures are idle between 40-45C and high load up to 75-80C, which I think is pretty reasonable on the stock cooler.

Today I went out and bought a Corsair CX650M. No change, the system still powers off suddenly. I am going to keep the new PSU since it was on my to-do list to get a better unit anyway, but this is a bit troubling.

I have also noticed that it tends to happen after a few minutes of sustained write to the NVMe SSD (mainly while downloading games on Steam). It has never happened during gameplay or benchmarking, just while writing to the drive.
 
Today I went out and bought a Corsair CX650M. No change, the system still powers off suddenly. I am going to keep the new PSU since it was on my to-do list to get a better unit anyway, but this is a bit troubling.

I have also noticed that it tends to happen after a few minutes of sustained write to the NVMe SSD (mainly while downloading games on Steam). It has never happened during gameplay or benchmarking, just while writing to the drive.
Interesting. Unfortunately I am not super knowledgable when it comes to NVMe drives, however, that still strikes me odd that this happens when writing to the drive. This makes me suspect that there is something wrong with the motherboard, but I am not sure.
 
Solution

TRENDING THREADS