[Helios 300] CPU & GPU Temps Spike Up on External Monitor

Axiond

Hello all!

Helios 300, GTX 1060 6GB variant here.

The external monitor I use is AOC q3279vwfd8 (31'5" set on 2K resolution, 75hz).
I'm not sure if the temps (as shown below) are normal, considering I only use it for browsing. I've also done the undervolt using ThrottleStop, replacing the thermal paste and pads.

This only happens only when I connect to the external monitor (FYI: I only use the ext. monitor, the laptop monitor is off) - without ext. monitor, I get around 40c and GPU is off.

1. Is it possible to use the integrated CPU on the ext. monitor? I cannot find the setting anywhere.
2. What should I do to have better temps while using ext. monitor?
3. Should I get cooling pad?

Thank you so much in advance!

sri369

1. Is it possible to use the integrated CPU on the ext. monitor? I cannot find the setting anywhere.
--- Integrated GPU does not support external monitor.

If you haven't undervolted GPU, follow the links in my signature to accomplish the same.

I have the same 300 series with 1060 6GB and on external monitor via HDMI literally 100% of time, and my idle/low use temps are about 37-40C. My monitor is an Asus one with full HD display @ 60 Hz.

I did set my browser (I use waterfox) not use hardware rendering - simple reason, for movies and browsing CPU can handle the tasks. Plus this also helps when on battery mode on integrated screen where the 1060 doesn't even need to get powered on thereby increasing battery life.
So I would say check what programs use dedicated GPU instead of integrated GPU, and based on your necessity, force them use integrated GPU.

I have been on external monitor since like 20 mins now (post my gaming) and below is the screenshot from MSI. (I disable the graph, and enable it only when needed like this instance for example.)

Axiond

Thanks, buddy - however, it doesn't work. I saw a slight change in temp (around -1 to -2c) for some minutes and it went back again.
Basically, I use the same setup on TS and MSI AB as @Red-Sand a.k.a hotelhero - www.youtube.com/watch?v=nYW7ROFx2Kk

What I don't understand is though, the GPU's clocked to 1404mhz automatically when it's connected to ext. monitor

Red-Sand

Axiond said:

Thanks, buddy - however, it doesn't work. I saw a slight change in temp (around -1 to -2c) for some minutes and it went back again.
Basically, I use the same setup on TS and MSI AB as @Red-Sand a.k.a hotelhero - www.youtube.com/watch?v=nYW7ROFx2Kk

What I don't understand is though, the GPU's clocked to 1404mhz automatically when it's connected to ext. monitor

That's because it's an Optimus system i.e. dual GPUs. Normally the Nvidia GPU is idle until in use... when you connect via HDMI, that external monitor is dedicated to the Nvidia GPU solely making it active (that's why it clocks up) 1404mhz is the base clock speed of your Nvidia GPU.

sri369

Red-Sand said:

Axiond said:

Thanks, buddy - however, it doesn't work. I saw a slight change in temp (around -1 to -2c) for some minutes and it went back again.
Basically, I use the same setup on TS and MSI AB as @Red-Sand a.k.a hotelhero - www.youtube.com/watch?v=nYW7ROFx2Kk

What I don't understand is though, the GPU's clocked to 1404mhz automatically when it's connected to ext. monitor

That's because it's an Optimus system i.e. dual GPUs. Normally the Nvidia GPU is idle until in use... when you connect via HDMI, that external monitor is dedicated to the Nvidia GPU solely making it active (that's why it clocks up) 1404mhz is the base clock speed of your Nvidia GPU.

@Red-Sand
When you connect it switches to external GPU, but @Axiond clock is going in frenzy mode to max normal speed of 1404. Mine stays at base speed of 139 when connected to external, and stays there unless i am doing something that uses GPU.

See below, its at 139, pushes up when i open up chrome and work on it, and then back to 139 when i close it. Looks like @axiond windows is not optimized properly and windows store seems to be using GPU (one of the major hidden culprits).

Red-Sand

It depends on if he has applications running on his Nvidia GPU and what they are.

Axiond

Thanks, bros!
@sri369 : can you please share your windows optimization?
FYI: Even no program is running on the graphic card and the ext. monitor is set on either 60hz / 75hz, I still get the same clock, temps and RPM

sri369

@Axiond I would suggest this first --> set your external monitor to normal full HD instead of 4k and to 60 Hz, and see if your GPU is still spiking to 1404 all the time.
Windows optimization:
1. Disable superfetch (I think they changed its name now, i don't have the 1809 update yet, so do not know its name right on top of my head)
2. Disable windows search
3. Disable/remove cortana completely (this is pretty much unless you don't value privacy much, or you do not use it)
4. You can remove all windows store apps, keeping only those you need. Personally, I even removed store. Only downfall would be apps like calculator, but I could find a windows 7 calculator and installed it.
5. Check for an app like Tuneup Utilities and disable programs you don't use frequently. Once set up, they are enabled and disabled automatically as you open and close them.
6. I would say stop using Predator Sense. This uses significant CPU and CPU cycles to display that graph. I use TS for CPU and MSI Afterburner for GPU undervolting (links in my signature). Suspend graph plotting once you are comfortable with your setup to save on those few cycles too.
7. Personally, I disabled all telemetry services too; these may look small but these do work in background collecting information, and i'd like my system totally idle. This also includes disabling the nvidia telemetry services.

sri369

@Axiond I connected my laptop to my TV - both external display only mode and in extended display mode, with laptop on HD and TV on 4k.
It went higher levels when being used, but under idle it settles back at 139. See the graph below - both usage and temps.

Graph includes temperature, memory speed and usage of GPU. CPU is temp only and that's less than 40 indicating pretty much barely used CPU.

All I did on my laptop is:
1. Optimize windows
2. Use felt pads under feet to lift its height to about an inch to allow better air flow on the underside.
3. Undervolting CPU and GPU (GPU under volting graph is visible on the screenshot).
4. My browser is set to use CPU only.

Axiond

@sri369 I finally get the same clock as yours but I have to reset the graphic card (I use Custom Resolution Utility 1.4.1) everytime, I have no clue what's going on
@Red-Sand care to pitch in?

Red-Sand

Well it could be that the forced resolution not being the screens native setting is causing the issue. 

sri369

Axiond said:

@sri369 I finally get the same clock as yours but I have to reset the graphic card (I use Custom Resolution Utility 1.4.1) everytime, I have no clue what's going on

Try to use with external only mode.
Do NOT duplicate displays.
Let me know if this helps.

tobimaru

What is your Windows 10 Power Plan setting?

Switching between 'Balanced' and "High Performance' has this effect. In 'Balanced' plan it will cycle the power state of your PCI-E connections and put the GPU in it's lowest setting on idle.

However, in 'High Performance' plan it is powering PCI-E connections no matter what, and your GPU is at full power even on idle.

This is how my Nitro 5 reacts when changing power plan settings. It may also be tied into the nVidia drivers as there is a power setting within the nVidia Control Panel. I believe the Os power plan superceeds this, but am not certain.