Question Ryzen 5 3600 high temperatures

[KRZ.]

While running a big installer that unpacks a heavily compressed file I noticed that my PC crashed. I started investigating and I came to the conclusion that it was crashing because of high temperatures, but:
1) I have a Corsair 275R Airflow case, which has a very good airflow
2) the wiring is almost perfect, there is nothing that might block airflow
The only big downpoint is that I'm using the stock cooler, but in any case I don't think that the CPU should reach 100° or more even with the stock cooler.

Also I tried something "bad": i took off the back panel of the case and while the CPU temp was over 100° I put my finger on the backplate of the mobo where the CPU is installed and it was not hot, I could have kept my fingers there for hours before I could get burned.
Now my question is: could it be that the sensors are reading wrong? Shouldn't that specific part of the mobo, being SO close to the CPU be hot as hell, or am I wrong?

Some pics of hwinfo



Also it looks like the CPU doesn't throttle, which is weird..

 

[coercitiv]

Sensor reporting looks weird on more than just temps, power usage is also through the roof considering your cooler.

If you're overclocking and/or have enabled XMP in BIOS, return to stock and redo thermal testing.

 

[Martimus]

If the backplate isn't hot, but the CPU is reporting 100C+, I would verify that the cooler is mounted properly. Maybe you made a simple mistake of forgetting to remove the plastic on the cooler, or you are missing thermal paste. It seems the heat isn't getting transferred for some reason.

 

[alexruiz]

If the backplate isn't hot, but the CPU is reporting 100C+, I would verify that the cooler is mounted properly. Maybe you made a simple mistake of forgetting to remove the plastic on the cooler, or you are missing thermal paste. It seems the heat isn't getting transferred for some reason.

The stock wraith stealth doesn't have plastic, and it already has the paste pre-applied.
I would discard that.

I would verify that the screws holding the heatsink are tightened all the way.

 

[alexruiz]

While running a big installer that unpacks a heavily compressed file I noticed that my PC crashed. I started investigating and I came to the conclusion that it was crashing because of high temperatures, but:
1) I have a Corsair 275R Airflow case, which has a very good airflow
2) the wiring is almost perfect, there is nothing that might block airflow
The only big downpoint is that I'm using the stock cooler, but in any case I don't think that the CPU should reach 100° or more even with the stock cooler.

Also I tried something "bad": i took off the back panel of the case and while the CPU temp was over 100° I put my finger on the backplate of the mobo where the CPU is installed and it was not hot, I could have kept my fingers there for hours before I could get burned.
Now my question is: could it be that the sensors are reading wrong? Shouldn't that specific part of the mobo, being SO close to the CPU be hot as hell, or am I wrong?

Some pics of hwinfo

View attachment 32614 View attachment 32615

Also it looks like the CPU doesn't throttle, which is weird..

Simple question.
Are you overclocking?
If so, reset the CMOS, bring it to stock settings.

A 3600 doesn't pull more than 88W in torture tests, so the 107W power consumption is eye popping.

 

[KRZ.]

Sensor reporting looks weird on more than just temps, power usage is also through the roof considering your cooler.

If you're overclocking and/or have enabled XMP in BIOS, return to stock and redo thermal testing.

Simple question.
Are you overclocking?
If so, reset the CMOS, bring it to stock settings.

A 3600 doesn't pull more than 88W in torture tests, so the 107W power consumption is eye popping.

I have not overclocked, what could be wrong then?
edit: also I don't have xmp enabled, I have game boost enabled

 

[damian101]

I have not overclocked, what could be wrong then?
edit: also I don't have xmp enabled, I have game boost enabled

Is "game boost" an option in the BIOS settings? Could definitely be some simple all-core overclock.
After you've resolved this issue I would definitely enable your RAM's XMP profile. It normally makes a bigger performance difference than overclocking with almost no increase in power draw.

 

[KRZ.]

Okay so I removed and reinstalled the cooler, also performed a CMOS reset, now the CPU doesn't go over 85° under load, also VID and package power are at their regular levels, I have no idea what was wrong..

 

[richierich1212]

Good thing you've resolved your issue.

 

[UsandThem]

I have not overclocked, what could be wrong then?
edit: also I don't have xmp enabled, I have game boost enabled

Never use the preconfigured overclocking features in the BIOS (e.g. game boost).
Almost all of them over-volt your CPU, and can cause issues like this.

 

[curryloti]

Looks like TS is running both hwinfo and another hardware monitoring program, doing so can screw up sensor readings.

 

[coercitiv]

I have no idea what was wrong..

You were unintentionally overclocking, in the sense that the Game Boost "feature" lets the CPU boost as high as possible on all-core loads while eliminating any power and temperature limits.

The result was your CPU was running at 4.2Ghz & 100W+ under heavy loads, with high voltage and no throttling in place to keep temperatures in check. Instability was likely caused by this configuration as well.

 

[aigomorla]

The only big downpoint is that I'm using the stock cooler,

get a new cooler.... you can see my honest feedback on that "should be recycled into beer can" excuse of a cooler they box wit the processor here:

Announcement - Ryzen Wraith Spire Cooler is Horrible.

I personally never use any stock coolers, but my cousin wanted to save some cash. Figured how bad could it be... Its horrible.... The wraith max and prisms were fairly acceptable. but the spire is just a straight up waste of aluminum which is better used to make into beer cans. Why AMD, you...

forums.anandtech.com

 

[coercitiv]

get a new cooler.... you can see my honest feedback on that "should be recycled into beer can" excuse of a cooler they box wit the processor here:

The OP's problem is already solved by returning the BIOS to stock settings instead of the MCE equivalent for the AMD platform.

MSI engineers messed up their "Game Boost" feature by pushing all-core boost to 4.2Ghz @ 1.4V and completely disabling thermal based throttling which would start progressively lowering clocks after reaching a certain thermal threshold. (which is probably around 85C based on what the OP is reporting).

 

[aigomorla]

The OP's problem is already solved by returning the BIOS to stock settings instead of the MCE equivalent for the AMD platform.

MSI engineers messed up their "Game Boost" feature by pushing all-core boost to 4.2Ghz @ 1.4V and completely disabling thermal based throttling which would start progressively lowering clocks after reaching a certain thermal threshold. (which is probably around 85C based on what the OP is reporting).

yeah i read that, but he is still better off upgrading the sink.
Its a horrible sink.... when you prime load it, it makes you really sad.