Nova 15.6 Overheating CPU

[JosDea]

Its possible that you got really unlucky with your CPU silicon. I could be wrong about this but I read somewhere that to get lucky with the higher end Ryzen 3900+ CPUs you need to have two decent CCD to get lucky with the silicon. Have you thought about manually OCing each CCD through Ryzen master? Run CCD0 a tiny but underlocked and CCD1 possibly a tiny bit overclocked? I don't know if its a great idea but I had heard a few tech people suggesting to manual OC each individual core of the Ryzen CPU's since each core is different. Hopefully a replasement CPU will fix all of these

I would have to run ccd0 at 2.8ghz all core vs ccd1 running at the stock 3.6 GHz for them to run at the same powerdraw under full load.

For the time being I've set the windows CPU power limit to 99% this prevents boost behaviours so the CPU will only hit 3ghz both CCD. Of course ccd0 still runs hot and draws to much power but it doesn't reach thermal shut down during gaming.

Not ideal at all though.

Edit : Having to send machine back hopefully it's as simple as a CPU switch.

 

[NoNastyNick]

Hi Guys,

These laptops do run hot, as we all know. However, there are times where the temperature is running too hot and close to thermal throttle point (95c) for extended periods of time. This is most definitely not normal.

Before you go to further lengths like @JosDea has and use copper shims - please bare in mind that there are some risks involved here also.
In JosDea's case, the mounting profile from the heatsinks was not adequate; we honestly don't know the reasons as to why this is. I'm not saying that JosDea may have maniuplated the metal, but it's most definitely a possibility when removing the cooling system.

JosDea is right regarding the pressure required to remove the heatsink - if you pull too hard, you risk damaging the CPU and rendering it completely useless. Run / use your laptop a few minutes prior to taking it apart, so you can easily remove the heatsink.

You shouldn't have to underclock to reduce the temperatures of the CPU / GPU - Again, these laptops are designed to run hot; hence the need for such high RPM fans on both the CPU and GPU. Yes, some people can't bear the thought of the fans spinning at high RPM's, but remember - this is a trade off for having a desktop replacement.

Currently, the best Thermal Compound based on user-reviews is Thermal Grizzly Kryonaut. This thermal paste is pretty thick and spreads far easier. I highly recommend not to use spreaders when applying thermal compound to Ryzen (or any CPU) for that matter, and please bear in mind - Ryzen CPU's have cores towards the side of the CPU. This means, the old "pea sized amount," will not be sufficient for Ryzen processors.

The best method I have found for applying the thermal paste for Ryzen, is using the X method - With dots of thermal paste around.
Yes, using too much thermal paste won't harm the CPU - however it will inhibit the effectiveness and purpose of what thermal paste really is. Thermal paste is used for transferring the heat from the DIE / IHS to the cooling system and filling in microgaps during the production process. I may be telling you guys how to suck eggs - but you wouldn't believe how often this is overlooked. The most important thing is to not have trapped air pockets.

If you have to remove the heatisinks - do not force it. As stated above - these are mallable copper pipes and if they are bent in anyway shape or form - this will inhibit the operation of the piping.

So, to summarise - If you are experiencing extremely high temperatures (close to thermal throttle 95c,) you have either issues with the thermal paste (too much / too little,) inadequate mounting pressure or blocked fans. JoseDea's case regarding CCD0 I would state is a rare exception.

 

[MathewEvan449]

@SUICIDALCHICK3N

Thank for the information. Since the bios update and a final over-usage of thermal paste. as long as CCC is in not in Performance Mode and I'll easily get hours of usage and can play games for 3-4 hours without issues. No under clocking used. I suspect as I've not seen to many posts about issues with the Nova 15.6 laptop that was a small batch issue. Either way so far so good.

 

[JosDea]

Hi Guys,

These laptops do run hot, as we all know. However, there are times where the temperature is running too hot and close to thermal throttle point (95c) for extended periods of time. This is most definitely not normal.

Before you go to further lengths like @JosDea has and use copper shims - please bare in mind that there are some risks involved here also.
In JosDea's case, the mounting profile from the heatsinks was not adequate; we honestly don't know the reasons as to why this is. I'm not saying that JosDea may have maniuplated the metal, but it's most definitely a possibility when removing the cooling system.

JosDea is right regarding the pressure required to remove the heatsink - if you pull too hard, you risk damaging the CPU and rendering it completely useless. Run / use your laptop a few minutes prior to taking it apart, so you can easily remove the heatsink.

You shouldn't have to underclock to reduce the temperatures of the CPU / GPU - Again, these laptops are designed to run hot; hence the need for such high RPM fans on both the CPU and GPU. Yes, some people can't bear the thought of the fans spinning at high RPM's, but remember - this is a trade off for having a desktop replacement.

Currently, the best Thermal Compound based on user-reviews is Thermal Grizzly Kryonaut. This thermal paste is pretty thick and spreads far easier. I highly recommend not to use spreaders when applying thermal compound to Ryzen (or any CPU) for that matter, and please bear in mind - Ryzen CPU's have cores towards the side of the CPU. This means, the old "pea sized amount," will not be sufficient for Ryzen processors.

The best method I have found for applying the thermal paste for Ryzen, is using the X method - With dots of thermal paste around.
Yes, using too much thermal paste won't harm the CPU - however it will inhibit the effectiveness and purpose of what thermal paste really is. Thermal paste is used for transferring the heat from the DIE / IHS to the cooling system and filling in microgaps during the production process. I may be telling you guys how to suck eggs - but you wouldn't believe how often this is overlooked. The most important thing is to not have trapped air pockets.

If you have to remove the heatisinks - do not force it. As stated above - these are mallable copper pipes and if they are bent in anyway shape or form - this will inhibit the operation of the piping.

So, to summarise - If you are experiencing extremely high temperatures (close to thermal throttle 95c,) you have either issues with the thermal paste (too much / too little,) inadequate mounting pressure or blocked fans. JoseDea's case regarding CCD0 I would state is a rare exception.

I'm considering getting a vapour / heatpipe combo heatsink created which should help things a lot. Also missed opportunity for additional cooling in terms of surface area of the fin array. Need to familiarise myself with cad first though.

 

[FerrariVie]

Hey guys... In my case, I have a Ryzen 3 3100 and was expecting it to run really cool on this chassis, which is the case for full CPU+GPU load (both at 80 degrees max). I even tried overclocking the CPU for a few hours for testing purposes, and I was able to get it stable to 4.35Ghz with 1.3V (stock boost is 3.9Ghz). However, the temps did go to lower 90s that way. Still far from throttling, but the fans get crazy loud, so I'll never try it again, as it's not worth it.

I'm ok with those temps, but my main concern is about the idle temps. While doing literally nothing, it stays around 55 degrees and while browsing, goes to 60. I can get the temps down to 50 idle and 55 while browsing by applying underclock and undervolt (3Ghz and 0.875V, while stock base clock is 3.6Ghz). No noticeable performance loss for browsing and the fans are not audible at those temps (around 24% fan speed), so while I'm happy with that, I still thought it should run cooler at idle. What do you think about those temps?

Another concern I have is with the heat pipes.. the GPU side is discoloured and has a few stains, that for me looks like thermal paste. Then someone might have tried to clean it and it looks a bit brushed. Do you think that it might cause issues, either now or in a few years? (picture below)

 

[NoNastyNick]

Hey guys... In my case, I have a Ryzen 3 3100 and was expecting it to run really cool on this chassis, which is the case for full CPU+GPU load (both at 80 degrees max). I even tried overclocking the CPU for a few hours for testing purposes, and I was able to get it stable to 4.35Ghz with 1.3V (stock boost is 3.9Ghz). However, the temps did go to lower 90s that way. Still far from throttling, but the fans get crazy loud, so I'll never try it again, as it's not worth it.

I'm ok with those temps, but my main concern is about the idle temps. While doing literally nothing, it stays around 55 degrees and while browsing, goes to 60. I can get the temps down to 50 idle and 55 while browsing by applying underclock and undervolt (3Ghz and 0.875V, while stock base clock is 3.6Ghz). No noticeable performance loss for browsing and the fans are not audible at those temps (around 24% fan speed), so while I'm happy with that, I still thought it should run cooler at idle. What do you think about those temps?

Another concern I have is with the heat pipes.. the GPU side is discoloured and has a few stains, that for me looks like thermal paste. Then someone might have tried to clean it and it looks a bit brushed. Do you think that it might cause issues, either now or in a few years? (picture below)

View attachment 18448

Hi mate,
I hope you're well.

Copper discolours over time - nothing to be concerned about. With regards to your Idle temperatures, the Ryzen's do idle far hotter than others. I wouldn't be overly concerned in all fairness. What you can do, is use the "slider" under balanced power-plan setting and set to the far left. Alternatively, if you want cooler idles you can limit the CPU even further by selecting (If i remember correctly,) power saving feature either within CCC or Windows.

How are you reading your temperatures? The "fanSpeed Setting" application is reading the hottest core, whilst Ryzen Master is reading the average across all cores.

 

[FerrariVie]

How are you reading your temperatures? The "fanSpeed Setting" application is reading the hottest core, whilst Ryzen Master is reading the average across all cores.

I'm using zen states, as it has a notification icon that shows the CPU temperature, but I actually don't know which figure that it reports. I'll do a comparison between it and ryzen master to see if they're any different.

 

[NoNastyNick]

I'm using zen states, as it has a notification icon that shows the CPU temperature, but I actually don't know which figure that it reports. I'll do a comparison between it and ryzen master to see if they're any different.

ahh okay - I've not really used ZenStates if i'm honest although I probably should (I hear great reviews.)

 

[FerrariVie]

ahh okay - I've not really used ZenStates if i'm honest although I probably should (I hear great reviews.)

Tested it now: Zen states does report the same temperatures as Ryzen Master

 

[MathewEvan449]

So the previous thermal issue seems to have reappeared, Unable to run the laptop at full load it will overheat and turn off. Suspect this is a issue with my laptop. Guessing no one else started to have the issue of late.

 

[demon28]

So the previous thermal issue seems to have reappeared, Unable to run the laptop at full load it will overheat and turn off. Suspect this is a issue with my laptop. Guessing no one else started to have the issue of late

Thats weird! What temperatures are you seeing during full load and which CPU are you running?

 

[MathewEvan449]

Full load the CPU sits around 79c and periodically will spike to around 90c. After which the screen will go blank, fans will keep spinning and nothing happens until i turn it off and restart.
CPU: AMD Ryzen 9 3900
GPU: Nvidia GTX 2070

IT was confirmed i was missing a BIOS update and tech support sent me the files to run, for a period there was no issues over the last week or so issues seem to be resurfacing.

 

[FerrariVie]

CPU overheating should not cause crashes, it should be able to thermal throttle and avoid similar issues. And anyway, 90° is indeed hot, but for a 12 cores CPU inside a laptop, short spikes into 90° are not that bad. From what I know, Zen2 CPUs will start reducing clocks and voltages after 75-80° to avoid getting closer to the 95° max, so the closer you get to 95, the lower clocks and voltages you'll get. But again, it should not give you crashes.

Everything that I'm saying here changes if you're manually tuning the voltages though, because if your voltages are not high enough for your clocks, then the crashes will happen.

I'm not so sure about VRMs... maybe that's them overheating? Someone had any experience with overheating VRMs and what the consequences are? Can they cause crashes?

 

[SpyderTracks]

CPU overheating should not cause crashes, it should be able to thermal throttle and avoid similar issues. And anyway, 90° is indeed hot, but for a 12 cores CPU inside a laptop, short spikes into 90° are not that bad. From what I know, Zen2 CPUs will start reducing clocks and voltages after 75-80° to avoid getting closer to the 95° max, so the closer you get to 95, the lower clocks and voltages you'll get. But again, it should not give you crashes.

Everything that I'm saying here changes if you're manually tuning the voltages though, because if your voltages are not high enough for your clocks, then the crashes will happen.

I'm not so sure about VRMs... maybe that's them overheating? Someone had any experience with overheating VRMs and what the consequences are? Can they cause crashes?

Yes, overheating VRMs would lead to instability, but that's a modded B450 board with a 65W part, there should be absolutely no problem with the VRMs at all, it's only with the 125W parts that you have to consider stronger VRM support.

If the VRMs are overheating, then it points to a problem with the cooling solution or voltage output.

 

[MathewEvan449]

I know the thermal pads don't fully cover my VRMs. I'm not doing any manual tuning at all, everything is just stock. And sadly i've not noticed any thermal throttling happening. just a straight up crash/blackscreen. The laptop is of course very well ventilated.

 

[SpyderTracks]

I know the thermal pads don't fully cover my VRMs. I'm not doing any manual tuning at all, everything is just stock. And sadly i've not noticed any thermal throttling happening. just a straight up crash/blackscreen. The laptop is of course very well ventilated.

When the screen goes blank, shine a torch and look very closely at the screen to see if you can make out an image.

It’s possible it’s a backlight issue and the laptop is fine.

 

[FerrariVie]

When the screen goes blank, shine a torch and look very closely at the screen to see if you can make out an image.

It’s possible it’s a backlight issue and the laptop is fine.

He could also observe the sounds of the games, if they keep going on just stop as well

 

[MathewEvan449]

So to clarify is it not a backlight issue 100%. I have no need to carry out that test as I use the laptop with the lid closed and use external displays. and they go black simply cause laptop has stopped. It has all the hallmarks of overheating protection mode. I've order some new grizzly paste and thermal pads. 1.0mill, 1.5mill and 2.0mill as i'm not sure of the thickness and have no paste to be able to take the heatsink off and check safely. Ill repaste. make sure NVRAM for the GPU is correctly padded and let you know if it continues.

This is the first machine from PC Spec i've had issues with over the years, Suspect just a pad, thermal paste issue.

 

[demon28]

Have you checked your thermal pads? Are all of them there/making good contact ? There isn't much info on what thickness you need to use ,but I took measurements of my pads last time I opened my laptop and I can find the pictures for you if you like. Although not professional measurements they should be clear enough on what goes where. I am pretty sure you will get to use all 3 sizes of the pads you ordered. What I am not sure about is if you need 3mil ones, I have heard of users stacking 2x1.5mm, but its probably not a great idea in terms of heat transfer...You can try asking PCS for a service manual. I asked them for one a few months ago but they said they didnt have it for the NOVA

 

[MathewEvan449]

So thermal pads 2.0m are too thick. 1.0m that were used i think were too think and not making good/full contact, Replaced with Artic Thermals 1.5m pads (order 2.0, 1.5 and 1.0 to be sure). was able to crank a game up to full everything and not have screen go black.. Other symptoms i started getting was screens going blank and coming back, any GPU accelerated application having performance issues depending on task.

So after thermal pad + paste replace things are looking positive for the first time. Will update again on weekend after had a few days trying to play games and run benchmarks see if it has solved the issue finally for me.