Question CPU temp fluctuates between 38 and 50C. What to do?

Jul 30, 2019
26
0
40
Hi,

Did you apply the thermal paste correctly?? to me those temps are not overly alarming, but I would suggest checking to ensure your liquid cooling is installed correctly?

Cheers,
 
Jul 27, 2019
58
2
45
You should be happy! That's a great temp for a CPU. Anywhere between 0° and 55° on idle and 0 to 70° under load is a very safe temp for your CPU. But of course, it depends on your cooler. Some coolers may get warmer than others. My ryzen 5 2600 with stock cooler gets 48° when idle and 68° under load, while my friend's ryzen 7 3400 with watercooling gets 39° on idle and 62° under load.
 
Last edited:

Sanderguy777

Honorable
Apr 1, 2017
29
0
10,530
I used the stress software and got a score of 2414.xx. Package temp (couldn't find core specific) is 69C when CPU is 100% on all cores and 4.0ghz.
Are those decent temps given the load? Does humidity have anything to do with cooler performance?
 

Karadjgne

Titan
Ambassador
That's perfectly normal. Has nothing to do with your hardware and everything to do with Windows. Windows itself is almost never idle. It's constantly doing background stuff and everytime it opens a program, it creates a load. That load creates work on the cores, which creates heat. Heat travels from the cores, through the IHS, through the paste, to the cooler base, gets absorbed by the heatpipes, travels up the fins and finally gets dissipated. All that took several seconds. Your pc will read the temps in your cores 2x a second. There's no way to change that, thermal properties being what they are, your pc is checking temps far faster than the cooler can deal with them.

Cooler makes exactly no difference. You could have a stock cooler, giant NH-D15 or 360mm AIO. It'll still do it.

So you will always get temporary spikes from 30 to 50, unless you shut down all of windows background stuff, like Cortana, windows store, Antivirus, index, net checks, temp checks......
 
  • Like
Reactions: BringerOfTea

Sanderguy777

Honorable
Apr 1, 2017
29
0
10,530
That's perfectly normal. Has nothing to do with your hardware and everything to do with Windows. Windows itself is almost never idle. It's constantly doing background stuff and everytime it opens a program, it creates a load. That load creates work on the cores, which creates heat. Heat travels from the cores, through the IHS, through the paste, to the cooler base, gets absorbed by the heatpipes, travels up the fins and finally gets dissipated. All that took several seconds. Your pc will read the temps in your cores 2x a second. There's no way to change that, thermal properties being what they are, your pc is checking temps far faster than the cooler can deal with them.

Cooler makes exactly no difference. You could have a stock cooler, giant NH-D15 or 360mm AIO. It'll still do it.

So you will always get temporary spikes from 30 to 50, unless you shut down all of windows background stuff, like Cortana, windows store, Antivirus, index, net checks, temp checks......

Wow, I have never thought about that. I mean I knew that the cores were never REALLY idle, but I had no idea that them just working caused SO much heat.

Thank you guys for telling me that the $125 I just spent was in vain.... LOL I'm keeping it, but only because I am NOT tinkering with that thing again (putting it in was a PAIN) and because it works better than the warped stock heatsink.
 
Jul 27, 2019
58
2
45
By the way, I just did a UserBenchmark test and got an error code on my photo storage hdd.
https://www.userbenchmark.com/UserRun/18928878
The drive in question is the SEAGATE Barracuda 2tb. I don't know what a RAM cache is or why their about page says it can wipe data if powered off.... Please help, i know this question is out of the scope of this post, but I just figured I'd ask.
could be be because you have a sata 3.0 hdd and you have it plugged in to a sata 2.0 port
 

Karadjgne

Titan
Ambassador
Ram is temporary storage. It's got no provision for keeping any data after power loss the way an SSD or hdd does. Doesn't matter if it's vram on gpu, or system ram or ram cache on a hdd.

Ram cache is a holder. When data is read/written to the hdd, it's got to be held somewhere while the reading/writing is in progress. So this e cpu tells the hdd it needs file XYZ.gif, the hdd starts reading it storing it temporarily in the cache, and when done reading the whole file, ships it complete to the system ram, which holds it until the cpu is fully ready for it and has thread room and available bandwidth.

Data corruption happens when writing something new or changed to the hdd, it's in the middle of doing so when you get power loss. That wipes out the cache, but the hdd has only half the info written. That's common to any hdd/ssd, even one that's fully functional. If the cache is bunk, or not working correctly, sudden power loss can wipe out the entire file as there's no buffer between the ram and hdd.
 

Karadjgne

Titan
Ambassador
Sata3 equipment is backwards compatible with sata2 ports, it'll just be slow as molasses in Siberia, mid-winter. Shouldn't cause a cache error.

I'd run the smart check on the drive, chkdsk, crystal info etc, see if it's not a hardware error, like a bad drive. If it is, I'd back it up fully, then see about rma or warranty.
 

Sanderguy777

Honorable
Apr 1, 2017
29
0
10,530
I couldn't get chkdsk to work, but I got and used CrystalDisk. I got some numbers, but I don't know what they mean, or if they tell me what the problem is.

Seq: Read 229.4 Write 221.7
4KiB: 2.253 1.563
Q8T8
4KiB: 2.186 1.558
Q32T1
4KiB .762 1.635
Q1T1

I know the seq read write, but I have no idea what anything else means and I don't see a way to fix the RAM cache thing. The RAM is fine though, I think, because I'm using the computer to type this and it is working just fine as far as I know.
As far as ports, I think they are all the same on the mobo. They just say SATA and are all bunched together with only numbers like 1-8, no 2.0 or 3.0...
 
I expect to see 10-15c. over ambient when idle if your cooler is mounted well and is functioning.
In particular, a H100I should be at the lower end of that range.
38c. is ok, but if you really are at 50c when idle you may have an issue with your cooler.
HWinfo will give you the current, minimum and maximum temperatures.
It should also let you verify that the cpu cooler pump is spinning.
Another possibility is that your cooler has lost fluid or the circulation is obstructed.

3400G needs no more than the stock cooler so replacing the cooler with a cryorig H7 or similar air cooler with a 120mm fan would not be difficult.
 

Sanderguy777

Honorable
Apr 1, 2017
29
0
10,530
I expect to see 10-15c. over ambient when idle if your cooler is mounted well and is functioning.
In particular, a H100I should be at the lower end of that range.
38c. is ok, but if you really are at 50c when idle you may have an issue with your cooler.
HWinfo will give you the current, minimum and maximum temperatures.
It should also let you verify that the cpu cooler pump is spinning.
Another possibility is that your cooler has lost fluid or the circulation is obstructed.

3400G needs no more than the stock cooler so replacing the cooler with a cryorig H7 or similar air cooler with a 120mm fan would not be difficult.

I already replaced the stock cooler with the Corsair, and I think it is mounted correctly. If you think it IS mounted badly and will affect the actual life of the device, I might be able to take it in to my IT guy, but I'm really busy with school.

On the other hand, I do need help with the HDD issue that has cropped up....
 
I already replaced the stock cooler with the Corsair, and I think it is mounted correctly. If you think it IS mounted badly and will affect the actual life of the device, I might be able to take it in to my IT guy, but I'm really busy with school.

On the other hand, I do need help with the HDD issue that has cropped up....
Your post asked about temperatures, that is what I responded to.
I do not see any real problem with temperatures.

On the HDD issue, look to your software. Somewhere, it would appear that it has implemented a ram cache.
That is where a portion of your ram is dedicated to hold data from your HDD. If the data desired is resident in ram, then I/O is vastly faster for reads and writes. A potential issue comes when the pcx shuts down either voluntarily or abnormally.
In a normal shutdown the ram cache software will write out any updated data to the HDD. OTOH, if it is a crash, your HDD will not have what it should have, potentially causing problems.
CHKDSK is an app that looks for mismatched HDD links. But, it can not fix all problems.
Probably best to revisit how the HDD is managed and turn off the more aggressive options. It may be too late.
 

Sanderguy777

Honorable
Apr 1, 2017
29
0
10,530
Your post asked about temperatures, that is what I responded to.
I do not see any real problem with temperatures.

On the HDD issue, look to your software. Somewhere, it would appear that it has implemented a ram cache.
That is where a portion of your ram is dedicated to hold data from your HDD. If the data desired is resident in ram, then I/O is vastly faster for reads and writes. A potential issue comes when the pcx shuts down either voluntarily or abnormally.
In a normal shutdown the ram cache software will write out any updated data to the HDD. OTOH, if it is a crash, your HDD will not have what it should have, potentially causing problems.
CHKDSK is an app that looks for mismatched HDD links. But, it can not fix all problems.
Probably best to revisit how the HDD is managed and turn off the more aggressive options. It may be too late.

What do you mean by too late. AND where do I look for "aggressive options"?
I have anhc (I cant remember if that is right. The ssd setting) set in bios, but I was following a build video from Bitwit or Jay's 2 cents that said that was the setting to use (but they ONLY had ssd's, no HDDs).
 

Sanderguy777

Honorable
Apr 1, 2017
29
0
10,530
I AM going to take it to my IT guy. He should be able to fix it and then I can quit worrying about it.

And to answer your question, it DOES reach 56C. I just saw it do it yesterday after using Chrome and opening (not using) two photo editors.

Also, I'm sorry for the way I said I wanted help with the HDD. It was rude. I didn't mean for it to be, I am just impatient that my build is still having issues...
 
By aggressive options, I mean to turn off HDD cacheing.

By too late, I mean that your HDD may have corrupted files with no way to recover except from a backup.

As to temperatures, I would not worry unless you see something like 75c. under heavy load.
 

Sanderguy777

Honorable
Apr 1, 2017
29
0
10,530
By aggressive options, I mean to turn off HDD cacheing.

By too late, I mean that your HDD may have corrupted files with no way to recover except from a backup.

As to temperatures, I would not worry unless you see something like 75c. under heavy load.


Ok. Corruption isn't a problem since there aren't any files on it.
How does one turn ram cache off (I didnt turn it on that I know of).

I will try to run it hard tonight to see the temps. I have COD black ops 2 if that will run it hard enough. Otherwise I can run 2 photo editors with a few dozen layers...
 

Sanderguy777

Honorable
Apr 1, 2017
29
0
10,530
So, my IT guy fixed the problem. There turns out there was no problem. The software was seeing that the hybrid drive was an ssd AND a hdd. He said (as you guys did) that the heat was common.

It's working great now! Thank you guys!