CPU temperatures stay as high as 70°C even with 5% CPU usage

[Mintymandy34]

My system info:

Code: Select all

$ inxi -Fxz
System:
  Host: user-PC Kernel: 4.18.0-15-generic x86_64 bits: 64 compiler: gcc 
  v: 7.3.0 Desktop: Cinnamon 4.0.9 Distro: Linux Mint 19.1 Tessa 
  base: Ubuntu 18.04 bionic 
Machine:
  Type: Laptop System: HP product: OMEN by HP Laptop 15-dc0xxx v: N/A 
  serial: <filter> 
  Mobo: HP model: 84DA v: 93.21 serial: <filter> UEFI: AMI v: F.07 
  date: 10/30/2018 
Battery:
  ID-1: BAT0 charge: 47.9 Wh condition: 47.9/47.9 Wh (100%) 
  model: HP Primary status: Full 
CPU:
  Topology: 6-Core model: Intel Core i7-8750H bits: 64 type: MT MCP 
  arch: Kaby Lake rev: A L2 cache: 9216 KiB 
  flags: lm nx pae sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx bogomips: 52992 
  Speed: 800 MHz min/max: 800/4100 MHz Core speeds (MHz): 1: 800 2: 800 
  3: 800 4: 800 5: 800 6: 800 7: 800 8: 800 9: 800 10: 800 11: 800 12: 800 
Graphics:
  Device-1: Intel vendor: Hewlett-Packard driver: i915 v: kernel 
  bus ID: 00:02.0 
  Device-2: NVIDIA GP107M [GeForce GTX 1050 Ti Mobile] 
  vendor: Hewlett-Packard driver: nvidia v: 390.77 bus ID: 01:00.0 
  Display: x11 server: X.Org 1.19.6 driver: modesetting,nvidia 
  unloaded: fbdev,nouveau,vesa resolution: 1920x1080~144Hz 
  OpenGL: renderer: GeForce GTX 1050 Ti/PCIe/SSE2 v: 4.6.0 NVIDIA 390.77 
  direct render: Yes 
Audio:
  Device-1: Intel Cannon Lake PCH cAVS vendor: Hewlett-Packard 
  driver: snd_hda_intel v: kernel bus ID: 00:1f.3 
  Sound Server: ALSA v: k4.18.0-15-generic 
Network:
  Device-1: Intel Wireless-AC 9560 [Jefferson Peak] driver: iwlwifi 
  v: kernel port: 5000 bus ID: 00:14.3 
  IF: wlo1 state: up mac: <filter> 
  Device-2: Realtek RTL8111/8168/8411 PCI Express Gigabit Ethernet 
  vendor: Hewlett-Packard driver: r8169 v: 2.3LK-NAPI port: 3000 
  bus ID: 03:00.0 
  IF: eno1 state: down mac: <filter> 
Drives:
  Local Storage: total: 1.03 TiB used: 170.34 GiB (16.2%) 
  ID-1: /dev/nvme0n1 vendor: Samsung model: MZVLW128HEGR-000H1 
  size: 119.24 GiB 
  ID-2: /dev/sda vendor: HGST (Hitachi) model: HTS721010A9E630 
  size: 931.51 GiB 
RAID:
  Hardware-1: Intel 82801 Mobile SATA Controller [RAID mode] driver: ahci 
  v: 3.0 bus ID: 00:17.0 
Partition:
  ID-1: / size: 98.04 GiB used: 9.96 GiB (10.2%) fs: ext4 
  dev: /dev/nvme0n1p2 
  ID-2: /home size: 902.43 GiB used: 160.38 GiB (17.8%) fs: ext4 
  dev: /dev/sda1 
  ID-3: swap-1 size: 18.63 GiB used: 0 KiB (0.0%) fs: swap 
  dev: /dev/nvme0n1p3 
Sensors:
  System Temperatures: cpu: 74.0 C mobo: N/A gpu: nvidia temp: 49 C 
  Fan Speeds (RPM): N/A 
Info:
  Processes: 317 Uptime: 4h 28m Memory: 15.48 GiB used: 3.73 GiB (24.1%) 
  Init: systemd runlevel: 5 Compilers: gcc: 7.3.0 Shell: bash v: 4.4.19 
  inxi: 3.0.27 

I've been using Linux Mint for about a month now and I noticed the CPU temperatures remain quite high even in case of normal usage.
I've not seen it go below 60°C and it ramps up to 100°C on heavy usage.
80°C on normal usage, and I think it's quite high for the CPU.

Anything I could do help with this scenario.
Like detect what is causing this high temperature, logging temperatures or anything?
Anything that can bring the temperature under control.
Please help.

[all41]

wow--fairly new machine so cpu cooling path issues should not be of concern.
Perhaps the temperatures being reported are inaccurate--that would be my guess.
Intel chips will throttle back if certain core temps are exceeded. Do you notice any slowness?

Also do you still have Win on there? and how is it doing?

[thx-1138]

...you're not the only one:
https://h30434.www3.hp.com/t5/Notebook- ... -p/6952079
https://h30434.www3.hp.com/t5/Gaming-No ... -p/6814962
https://h30434.www3.hp.com/t5/Gaming-No ... -p/6991259
https://h30434.www3.hp.com/t5/Notebook- ... -p/6956285
https://h30434.www3.hp.com/t5/Gaming-No ... -p/7010202
https://h30434.www3.hp.com/t5/Gaming-No ... -p/7008223
And...
https://www.reddit.com/r/techsupport/co ... _problems/
https://www.reddit.com/r/pcgamingtechsu ... _problems/

...if you want a 100% honest answer: complain to HP, and hope they'll release a BIOS update that fixes it's behavior.
However, being a 6-Core gaming laptop, personally i'd certainly expect it to get relatively heated in the first place.

If you want somewhat of a partial diy workaround: use / configure TLP,
and if still not satisfied, then possibly further undervolt it. And...
Processes: 317 Uptime: 4h 28m Memory: 15.48 GiB used: 3.73 GiB (24.1%)
For comparison here...
Processes: 157 Uptime: 4h 36m Memory: 7.65 GiB used: 845.6 MiB (10.8%)
cpu temp is at 30 C & gpu temp at 35 C here at this moment.
You can absolutely disregard that, as that would be comparing apples with oranges (it's a dual core i5-6200).
Granted, i'm quite likely guilty of having a relatively spartan computing life...but still:
twice the number of processes?...
I'd certainly try to lighten up the workload.
Not on Cinnamon myself though, maybe others can comment on that.

[Mintymandy34]

wow--fairly new machine so cpu cooling path issues should not be of concern.
Perhaps the temperatures being reported are inaccurate--that would be my guess.
Intel chips will throttle back if certain core temps are exceeded. Do you notice any slowness?

Yes, when the temperatures go beyond 80°C, I notice lags.

Also do you still have Win on there? and how is it doing?

Nah, I got rid of Windows totally.
I've tried other OS, Linux Mint Xfce, Lubuntu and Puppy Linux from live usb and Lubuntu couldn't go beyond 60°C on medium usage.
Mint Xfce started around 50°C and settled at 65°C.

Any ideas about anything more I could do?

[Mintymandy34]

...you're not the only one:
https://h30434.www3.hp.com/t5/Notebook- ... -p/6952079
https://h30434.www3.hp.com/t5/Gaming-No ... -p/6814962
https://h30434.www3.hp.com/t5/Gaming-No ... -p/6991259
https://h30434.www3.hp.com/t5/Notebook- ... -p/6956285
https://h30434.www3.hp.com/t5/Gaming-No ... -p/7010202
https://h30434.www3.hp.com/t5/Gaming-No ... -p/7008223
And...
https://www.reddit.com/r/techsupport/co ... _problems/
https://www.reddit.com/r/pcgamingtechsu ... _problems/

...if you want a 100% honest answer: complain to HP, and hope they'll release a BIOS update that fixes it's behavior.
However, being a 6-Core gaming laptop, personally i'd certainly expect it to get relatively heated in the first place.

My bad, I learnt how bad HP support really is after moving to Linux.
I can complain about it, but I think it's highly unlikely that they'll release anything.
I did about this before purchasing the laptop and forums said updating to newer BIOS will resolve the issue but it seems like it didn't.
Another solution was to use Intel XTU for undervolting but I don't know how to use it in Linux.
Are there any proper Linux equivalent to Intel XTU?

If you want somewhat of a partial diy workaround: use / configure TLP,
and if still not satisfied, then possibly further undervolt it. And...
Processes: 317 Uptime: 4h 28m Memory: 15.48 GiB used: 3.73 GiB (24.1%)
For comparison here...
Processes: 157 Uptime: 4h 36m Memory: 7.65 GiB used: 845.6 MiB (10.8%)
cpu temp is at 30 C & gpu temp at 35 C here at this moment.
You can absolutely disregard that, as that would be comparing apples with oranges (it's a dual core i5-6200).
Granted, i'm quite likely guilty of having a relatively spartan computing life...but still:
twice the number of processes?...
I'd certainly try to lighten up the workload.
Not on Cinnamon myself though, maybe others can comment on that.

Will read about tlp.
I've tlp but the person who suggested it to me recommended to leave it at default settings.
And is there any frontend for tlp?
And about the processes, I'll check if that's an issue in Cinnamon.
Are you on Xfce?

Thanks for helping out.
I'll post back details from the live usb experiences of the other distros for comparison.
And is there a way to generate a report of temperature usage or CPU usage?

[philotux]

For temperature:
install lm-sensors and run it in a terminal with the command sensors. You could monitor cpu-usage and temperature graphically in real time withpsensor

psensor.png

For cpu usage, memory usage etc.top orhtopin a terminal:

htop.png

[sleeper12]

I don't know if this is any help to you, but my Dell Latitude E6410 was overheating & shutting down. Tried everything, nothing worked, even 2 repair shops had no clue. Before throwing it out the window , I put copper shims on the CPU/GPU with thermal paste on both sides. Covered the GPU with a thermal pad (more paste) &, much to my surprise, it actually worked. I'm posting on that machine now. So, something to consider, if all else fails.

Edit: It was the copper shims & thermal pad that made the difference. Thermal paste alone wasn't enough in this case.

[Mintymandy34]

For temperature:
install lm-sensors and run it in a terminal with the command sensors. You could monitor cpu-usage and temperature graphically in real time withpsensor
psensor.png

For cpu usage, memory usage etc.top orhtopin a terminal:


htop.png

Thanks for suggesting them, I'm already using them.
However I want something that could be used for logging data, like logging for hours so I could compare them statistically for different distros and hence rule out the question of whether Cinnamon is causing this.
Monitoring helps, but I can't get the overall data charts or values, not good for comparative study.
If either of those solutions you've suggested, have the automatic logging feature(that I don't know as of yet), let me know, I'll use them.
Though I would like an additional tool, meant for logging temperature data, usage data.

Thanks for your reply.

[Mintymandy34]

I don't know if this is any help to you, but my Dell Latitude E6410 was overheating & shutting down. Tried everything, nothing worked, even 2 repair shops had no clue. Before throwing it out the window , I put copper shims on the CPU/GPU with thermal paste on both sides. Covered the GPU with a thermal pad (more paste) &, much to my surprise, it actually worked. I'm posting on that machine now. So, something to consider, if all else fails.

Yes, thermal paste is also a huge consideration.
There were 3 known possible solutions to this, bios update didn't fix anything, I've no idea how to do undervolting in Linux and third one was applying thermal pastes.
HP has provided with some, but in every review that I've read, people said, "Get rid of it, and paste a better one".
Could you please provide me with some more info on that, like what products you chose, methods or articles that helped you through that process?
I'm a noob, , so I don't know much about the different types of pastes, so please guide me right.

And thanks for this, I'll surely go this way if overclocking/undervolting or any other simpler solution fails.
It would be really nice if you could put the temperature data for me to compare.

[powerwagon75]

.....If either of those solutions you've suggested, have the automatic logging feature(that I don't know as of yet), let me know, I'll use them...

Psensor has a very basic logging feature, under Psensor Preferences, the Sensors tab. You got my curiosity going, and thought I'd try using it once during this session of Folding. Data is pretty bare, but is somewhat readable. Under that Sensors tab, check the box to enable logging. I experimented with a 5 "second" interval, and a Sensors logging interval of 9000 seconds. (I presumed this time interval to be the amount of time the log would run) It would appear that the "seconds" should really be "minutes", as the log is being updated every 5 minutes. Also, it appears that the maximum logging interval is 6500 seconds, or at least that is the maximum number it will show you, if you navigate back to the Sensors tab. I went in to dconf Editor and changed it to 18000. (50 hours) Will have to wait and see if it actually honors that value.
Lines 22~28 show the timed interval data captures..somewhat inconsistent, as it seems to miss the GPU fan speed randomly.

psensor-log.png

Reading across the entry it seems to log CPU core temps, GPU temp, GPU utilization percent, fan speed, hdd temps, and CPU usage, free memory. However, it seems incapable of grabbing all the data at each interval..something is usually missing.

It's probably not exactly what you were looking for, but you can copy the comma-separated values out and drop in to a Libre spreadsheet and chart/graph it as you wish. (make sure to check the box for separating by commas when dropping the data in.)

Log file is in your home folder, inside the .psensor folder.

[trytip]

don't bother with thermal paste your laptop is way too new to have an issue. i would suggest downloading a new distro that uses a current kernel and test it live. check temperatures there

try the new http://www.mageia.org/en/downloads/prerelease/ even if it's beta it's only for comparing and testing. make the usb the same way you would linux
select live media and choose a desktop. if you're not familiar with plasma or gnome xfce is the most natural choice

[sleeper12]

Is this Bios update new to you?
https://www.youtube.com/watch?v=HeMRUFSb0d0

If your laptop is still under warranty, opening it up will void the warranty, so be careful.
I'm no expert, but any arctic silver paste is ok far as I know. Google around & check Youtube for 'How to' info.
As for temps, well my laptop temps always said normal, yet it would shutdown after a minute or less from overheating. Go figure. Right now, CPU 47, GPU 57.

[Mintymandy34]

.....If either of those solutions you've suggested, have the automatic logging feature(that I don't know as of yet), let me know, I'll use them...

Psensor has a very basic logging feature, under Psensor Preferences, the Sensors tab. You got my curiosity going, and thought I'd try using it once during this session of Folding. Data is pretty bare, but is somewhat readable. Under that Sensors tab, check the box to enable logging. I experimented with a 5 "second" interval, and a Sensors logging interval of 9000 seconds. (I presumed this time interval to be the amount of time the log would run) It would appear that the "seconds" should really be "minutes", as the log is being updated every 5 minutes. Also, it appears that the maximum logging interval is 6500 seconds, or at least that is the maximum number it will show you, if you navigate back to the Sensors tab. I went in to dconf Editor and changed it to 18000. (50 hours) Will have to wait and see if it actually honors that value.
Lines 22~28 show the timed interval data captures..somewhat inconsistent, as it seems to miss the GPU fan speed randomly.
psensor-log.png
Reading across the entry it seems to log CPU core temps, GPU temp, GPU utilization percent, fan speed, hdd temps, and CPU usage, free memory. However, it seems incapable of grabbing all the data at each interval..something is usually missing.

It's probably not exactly what you were looking for, but you can copy the comma-separated values out and drop in to a Libre spreadsheet and chart/graph it as you wish. (make sure to check the box for separating by commas when dropping the data in.)

Log file is in your home folder, inside the .psensor folder.

Thanks, that's a huge help.
Seriously that's really huge in the context.
I'm logging the data, will plot it in Libre Calc.
I'm running a 60 minutes test, I'm planning to capture data from the default graph from psensor, I'll capture the logging data along with it for verification/comparison.

Thanks a lot.

[Mintymandy34]

don't bother with thermal paste your laptop is way too new to have an issue. i would suggest downloading a new distro that uses a current kernel and test it live. check temperatures there

try the new http://www.mageia.org/en/downloads/prerelease/ even if it's beta it's only for comparing and testing. make the usb the same way you would linux
select live media and choose a desktop. if you're not familiar with plasma or gnome xfce is the most natural choice

That's the last resort, I've 3 years warranty, so I'll try go with the simpler ways.
And thanks for suggesting the distro.
I've already tested, Mint 19.1 Cinnamon, Xfce, Ubuntu Studio 18.10, Kubuntu 18.10, Ubuntu Mate 18.10, Lubuntu 18.04.1 and Xenialpup 7.5 in live USB environment.
But I suffered a lot of ACPI exceptions and I couldn't suspend or reboot quite easily, so I couldn't run extensive tests like rendering and all.
I only tested for idle usage and surfing the web and media contents.
Will try Mageia 7.
Thanks for the suggestion.

[Mintymandy34]

Is this Bios update new to you?
https://www.youtube.com/watch?v=HeMRUFSb0d0

Yes, it is.

If your laptop is still under warranty, opening it up will void the warranty, so be careful.
I'm no expert, but any arctic silver paste is ok far as I know. Google around & check Youtube for 'How to' info.
As for temps, well my laptop temps always said normal, yet it would shutdown after a minute or less from overheating. Go figure. Right now, CPU 47, GPU 57.

Yup, mine isn't shutting down or anything, but it's pretty hot even after 30 minutes of simple tasks like browsing internet and media files.
Rendering, oh well, in 5 minutes, it spikes up beyond 95 degrees and I've seen 100 at a max but it's way too hot for my liking.
And about paste, I'm not going to apply it myself, I'll try convincing HP support to do it (not sure if they'll, they took 2 hours to add a RAM stick, so).
Paste is a last resort, like really at the bottom of the barrel, it's a strong resort but not the viable one currently.

[Acewiza]

If that laptop is under warranty it needs to go back. There is undoubtedly some hardware defect in the cooling system. My guess would be a broken fastener somewhere.

[sleeper12]

Ok under warranty, so have HP fix it rather than the headaches of trying to do it yourself.

[Mintymandy34]

Ok under warranty, so have HP fix it rather than the headaches of trying to do it yourself.

Yeah, I checked for the BIOS update, I've the newer one already installed.
About warranty, to be honest, if I send it now, it's not coming back in two months or so, or maybe more, I've had worse experiences in the past.
I've a lot of work coming up next week, so it's getting tight.

I wouldn't want to send them this laptop unless it is absolutely beyond any resolution.
I hope you understand.
And I'm sorry about this.
I hope to have a better experience this time around.

[Mintymandy34]

.....If either of those solutions you've suggested, have the automatic logging feature(that I don't know as of yet), let me know, I'll use them...

Psensor has a very basic logging feature, under Psensor Preferences, the Sensors tab. You got my curiosity going, and thought I'd try using it once during this session of Folding. Data is pretty bare, but is somewhat readable. Under that Sensors tab, check the box to enable logging. I experimented with a 5 "second" interval, and a Sensors logging interval of 9000 seconds. (I presumed this time interval to be the amount of time the log would run) It would appear that the "seconds" should really be "minutes", as the log is being updated every 5 minutes. Also, it appears that the maximum logging interval is 6500 seconds, or at least that is the maximum number it will show you, if you navigate back to the Sensors tab. I went in to dconf Editor and changed it to 18000. (50 hours) Will have to wait and see if it actually honors that value.
Lines 22~28 show the timed interval data captures..somewhat inconsistent, as it seems to miss the GPU fan speed randomly.
psensor-log.png
Reading across the entry it seems to log CPU core temps, GPU temp, GPU utilization percent, fan speed, hdd temps, and CPU usage, free memory. However, it seems incapable of grabbing all the data at each interval..something is usually missing.

It's probably not exactly what you were looking for, but you can copy the comma-separated values out and drop in to a Libre spreadsheet and chart/graph it as you wish. (make sure to check the box for separating by commas when dropping the data in.)

Log file is in your home folder, inside the .psensor folder.

Couldn't understand the log file properly.
And that file updates after 5 minutes, and I've set the Sensors Logging Interval value to be 300 seconds.
And I've Measure Update Interval at 2 seconds, can't understand what that changes.


Click on the image for a larger view.

The first section is Idle usage, second, third and fourth are normal usage, surfing internet and file browsing and all, the last one is idle usage.
And now:

Code: Select all

 sensors
coretemp-isa-0000
Adapter: ISA adapter
Package id 0:  +49.0°C  (high = +100.0°C, crit = +100.0°C)
Core 0:        +47.0°C  (high = +100.0°C, crit = +100.0°C)
Core 1:        +47.0°C  (high = +100.0°C, crit = +100.0°C)
Core 2:        +46.0°C  (high = +100.0°C, crit = +100.0°C)
Core 3:        +47.0°C  (high = +100.0°C, crit = +100.0°C)
Core 4:        +45.0°C  (high = +100.0°C, crit = +100.0°C)
Core 5:        +49.0°C  (high = +100.0°C, crit = +100.0°C)

acpitz-virtual-0
Adapter: Virtual device
temp1:        +48.0°C  

iwlwifi-virtual-0
Adapter: Virtual device
temp1:        +44.0°C  

pch_cannonlake-virtual-0
Adapter: Virtual device
temp1:        +51.0°C  

I'm browsing Firefox, 5 tabs and am slightly above normal usage.

And I read on HP forums that Core 0 to Core 5 are actual temperature of cores and pch_cannonlake-virtual-0 is not subjected to the processor cooling.
This is the value that is spiking off, I'm not quite sure what all these really mean, so not sure what's the problem.
I'll try the suggested distro along with the others, this time around with proper logging and all.
Thanks for helping out.

[HaveaMint]

Where you have the laptop sitting when your using it doesn't restrict the airflow an anyway does it?