Core multiplier dropping mid stress test, followed by complete lock up

PlatyCS

Reputable
Jul 17, 2017
10
0
4,510
Hey all, I'll start off my listing my current specs before addressing the problem:

CPU - AMD FX-8350
CPU Fan - CoolerMaster Hyper 212 EVO w/ 2 fans
Mobo - ASUS M5A97 LE R2.0
GPU - GTX 980ti 6GB (MSI Golden Edition)
Ram - 8gb DDR3 1600MHz
PSU - Corsair CX600

The problem I'd like to address is specific to when I run a prime-95 small FFT test after having modified my core clock as little as up to 4.2GHz from a default of 4.0GHz w/ 1.375VID. What happens, is the test will be going fine, temps never getting above 50C, and out of nowhere the core clock will drop to 1.4GHz, followed by the VID dropping to roughly 0.9V. I've also attempted also upping the VID temporarily to 1.4V just to see if the problem persists, and it does. Shortly after the clock and VID drop, if I don't stop the test promptly, the entire system will lock up, and the only way to do anything is by holding the power button to shut off the computer and rebooting.

My best guess would have to be there is some sort of BIOS setting that I have forgot to disable, or have not changed properly. However there's also the chance that there is something faulty with my hardware, which would be a less fun problem. Either way, I'd greatly appreciate some help in troubleshooting this problem and attempting to get my overclock going.

Thank you for your time
 
Solution
Most likely the vrms on your motherboard are overheating and slowing the CPU down to keep from melting. Open your side panel and aim a fan at the capicitors nearest the CPU and see if the issue goes away. Those boards aren't known for the ability to provide power to those CPUs.

Supahos

Expert
Ambassador
Most likely the vrms on your motherboard are overheating and slowing the CPU down to keep from melting. Open your side panel and aim a fan at the capicitors nearest the CPU and see if the issue goes away. Those boards aren't known for the ability to provide power to those CPUs.
 
Solution

PlatyCS

Reputable
Jul 17, 2017
10
0
4,510


First and foremost thank you for the lightning fast reply!

Seeing as my temperatures on my CPU weren't getting too high I took the liberty of removing one of my two 120mm fans off my heat sink and pointed it at the capacitors you mentioned. The immediate problem mentioned in my post has disappeared.

As I said, before, the clock would drop within the first 30 seconds of the test. Prime95 ran for roughly 5 minutes without this happening, although one of my cores did fail at around the 4 minute mark due to "hardware failure". I feel like this is no longer the concern of this post, so unless you're willing to continue to help troubleshoot if there is in fact a hardware malfunction in my CPU, I'll be on my merry way :)
 

PlatyCS

Reputable
Jul 17, 2017
10
0
4,510
In between when I added my reply and just now, I've upped the voltage back to 1.4V and have had p95 small FFT running with no hiccups whatsoever. CPU sitting at a comfortable 45C under full load.

Are you still positive the motherboard is iffy? It seems to be running quite smoothly now.
 

Supahos

Expert
Ambassador
It's probably fine at stock speeds it's not a horrible board, just not one you want to oc with. Everytime it gets hot enough to throttle you risk popping a capicitor and ruining the board. I'd go with what you have now and stop the prime tests
 

PlatyCS

Reputable
Jul 17, 2017
10
0
4,510
Sounds good to me. I was able to bump my cinebench score from a 535 to a 650 just from the 200MHz OC, so I won't get any more greedy. Thanks for all the help!