Curious....

[AvzNut75]

Does this seem strange to anyone else?

 

[trents]

Use AMDOverdrive instead to monitor temps. Some of those readings from HWMonitor Pro are obviously incorrect. The package temp looks about right, however. Are those max temps under sustained full load? Say Prime95 for 20 minutes?

 

[AvzNut75]

No...no, those are mostly idle temps. Out of habit, I am disinclined to use anything other than HWMonitor. I only have AISuite2 installed to monitor VRM temps. I had to glue some aluminum heatsinks that were laying around onto the VRM backplate to get the temps i'm getting

 

[Rainwater]

So were you idling at 95% cpu usage then too? 106c on Mainboard just seems Outlandish.

 

[trents]

AISuite is famous for causing problems.

Sometimes old habits block solutions. You need to confirm those high temp spikes are actually happening. If they actually happened and happened at idle something is using a lot of CPU slices in the background. Did you check processes in taskmanager to identify what is causing that? What about startup?

 

[Kenrou]

HWiNFO64 has been the most reliable for me so far - https://www.hwinfo.com/download.php

 

[AvzNut75]

So were you idling at 95% cpu usage then too? 106c on Mainboard just seems Outlandish.

Well, no....I wasn't fully idle the whole time. Panda ran a scan. However, you are right....it -does- seem outlandish.....

 

[AvzNut75]

AISuite is famous for causing problems.

Sometimes old habits block solutions. You need to confirm those high temp spikes are actually happening. If they actually happened and happened at idle something is using a lot of CPU slices in the background. Did you check processes in taskmanager to identify what is causing that? What about startup?

I only use AISuite because, to my knowledge, it is the only app that will monitor my VRM temps. The discrepancy in temps/volts/RPMs seems entirely random, and not tied to any particular usage. I have used TM to try to determine where the spikes are originating, but have not found any connection thus far. Frankly, the fact that im able to type this post (combined with the value extremes) leads me to believe that none of the readings are accurate. One thing....when I am running Prime95, the strange spikes disappear.....

- - - Updated - - -

Even stranger.....I have gamed Dragon Age Inquisition for a few hours @ 5 Ghz without any problems....

 

[trents]

Have you confirmed that the temp spikes coincide with CPU activity spikes?

Is there a malware process going on here?

 

[Soulcatcher668]

AI suite temps are not to be trusted.
I do believe you need to some proper testing on this overclock. If your board is indeed at 106, it is not long for this world.

 

[kyfire]

Funny how everyone keys in on the 106 reading but seemingly missed the -117 min reading while the CPU min was 18.5. CPUID is known to give false readings, especially on AMD FX chips.

 

[RJARRRPCGP]

Funny how everyone keys in on the 106 reading but seemingly missed the -117 min reading while the CPU min was 18.5. CPUID is known to give false readings, especially on AMD FX chips.

I'm not so sure about FX systems, but with FM2+, it's the norm to get a reading like that!

 

[Soulcatcher668]

Funny how everyone keys in on the 106 reading but seemingly missed the -117 min reading while the CPU min was 18.5. CPUID is known to give false readings, especially on AMD FX chips.

Ridiculously low reading are not abnormal but I was unaware of anomalous high readings.

It just seems that something is off here. Not and AMD guy so I will bow out.

 

[Alaric]

Use AMDOverdrive instead to monitor temps. Some of those readings from HWMonitor Pro are obviously incorrect. The package temp looks about right, however. Are those max temps under sustained full load? Say Prime95 for 20 minutes?

Tgis is what I would try for temps. I never had a problem with HWMonitor on my FX, but idle/ low load temp readings on the FX chips has always been off. It's the nature of the beast. The package temp has been all over the board, too. For an FX try AMDOverdrive and just check the temp under load. Idle temps are a mystery on the FX. Fortunately, unless you have a major hardware/cooling problem they don't matter.

 

[AvzNut75]

Here is what AOD says.....



- - - Updated - - -

Tgis is what I would try for temps. I never had a problem with HWMonitor on my FX, but idle/ low load temp readings on the FX chips has always been off. It's the nature of the beast. The package temp has been all over the board, too. For an FX try AMDOverdrive and just check the temp under load. Idle temps are a mystery on the FX. Fortunately, unless you have a major hardware/cooling problem they don't matter.

A search for tgis turned up nothing but catering software?

 

[trents]

Well, AMDOD didn't help either, did it? Thanks for trying. I would sure like to see those readings with AI Suite uninstalled.

 

[Alaric]

That doesn't look bad for an FX. The negative numbers are to be expected. There is an issue with FX temp sensor readings at anything lower than under load. RGone covered it in a post (or two) a while back. It would be nice if Zen addressed this, but Zen already promises to right a lot of wrongs already. Temp sensors may be a bit much to ask. LOL

 

[bob4933]

All your cores being EXACTLY the same temperature down to a hundreth of a degree seems unrealistic as well.

 

[AvzNut75]

Well, thank the Great Spirit for Panda Cloud Cleaner. It found the malware that regular scans had missed. Also, I installed HWiNFO64....much more accurate temps. Thanx fer the input erebody.

 

[Mr.Scott]

All your cores being EXACTLY the same temperature down to a hundreth of a degree seems unrealistic as well.

That's because there is only 1 temp sensor, not 1 for each core.