Need some help

[zel87]

Hello all. I am a complete novice when it comes to PC Modding/building and Overclocking. I have recently been doing a few cheap upgrades on my faithful old pre-built gaming system, mostly as practice because I eventually want to upgrade it to a budget Ryzen system when finances permit. Specs as follows:

CPU: (FX-4300) now AMD FX-6300

CPU Fan/Cooler setup: Noctua NH-U12S CPU Cooler with NF-F12 120mm Fan (plus another NF-F12 on back of case as exhaust fan)

GPU: (GTX 950 2GB) now 4095MB NVIDIA GeForce GTX 970 (Gigabyte) 'Windforce'

MOBO: MSI 760GM-P23(FX) (MS-7641)

RAM: 8GB DDR3 1866MHz

PSU: Cooler Master 550w 80+ White Non-Modular

Case: NZXT S210 (Large, well ventilated)

OS: Windows 10 Home 64-bit

The issue: I have been attempting to OC the CPU using AMD Overdrive. Nothing too adventurous as I have read that this MOBO is a bit rubbish? However, I am getting random extremely high temperature spikes jumping instantly from mid 40s to 200c+ when I run a stress test using RealBench 2.56 (I've been using Core Temp and Speccy to monitor temps). The spike happened at different times during the test, but both times after it had been running for at least a few minutes. This was at 4.2ghz and 1.365V with Turboboost disabled. It passed the stress test originally at these settings but then the system crashed during normal use. I tried playing around with a few things to no avail, went back to these settings and ran the stress test again and got the high temp spike. The case and all components (including heatsink and fans) were thoroughly cleaned before the new CPU and GPU went in, I also applied new thermal paste when installing the new CPU although it was cheap no-brand stuff off ebay.

Any thoughts on the issue would be welcome. Is it just a bad chip, have I damaged it, or both? Would better thermal paste make any difference? I have returned the settings to default and ran another stress test which it passed no problem (Max temp 40c).

 

[ninjag]

200°C?...that's got to be an error otherwise I'd be slapping a ribeye on it.

 

[Drumroll]

Clearly some software issue. The axPU would be gone now if ever close to 200C.
Thermal throttling would kick in, unless all safety issues are turned off, but still, no chance the CPU would still works.
I would recommend updating BIOS and software to get a reliable reading.

 

[modd1uk]

I use HW info for monitoring, I find it more reliable than other software...might need to disable a few settings though as it shows a lot.

 

[zel87]

Clearly some software issue. The axPU would be gone now if ever close to 200C.
Thermal throttling would kick in, unless all safety issues are turned off, but still, no chance the CPU would still works.
I would recommend updating BIOS and software to get a reliable reading.

Ok thanks. I checked the dates and I'm pretty sure I have the latest bios update but I will download the latest one off the manufacturers site and try it, see if it makes any difference.

I use HW info for monitoring, I find it more reliable than other software...might need to disable a few settings though as it shows a lot.

Ok, I'll give it a go. Thanks.

 

[TrojanWhore]

when you say 200+ degrees, is it specifically 256?

mine occasionally reads that in core temp - 256 is 2^8, i.e. the max integer value possible for an 8 bit register so I assume it is just a sensor issue and ignore it. (I've not seen the issue in HWmonitor or HWINFO64 yet that I rememeber)

When under a consistent load it stays within reason (still hot but ryzen does run hot), but transitions from idle to load, such as launching a game or loading between maps seems to trigger it.