Overheating issues

[CraigBarrett215]

I’m glad you like it, I think a lot of previous negative reports have been from I configured systems where temps will be high until windows is properly configured. They’re definitely not reports on properly working systems.

Doesn't that imply that PCS is not configuring systems correctly? My Recoil IV runs really hot - the keyboard is getting to around 45 degrees C when gaming and the area above the keyboard hits about 53 degrees C. I haven't checked the temperature of the bottom, but it's way too hot to have the laptop sitting on my lap - and it was configured by PCS. The reports of high running temperatures is something that put me off the Vyper, but it's a problem on the Recoil as well.

 

[SpyderTracks]

Doesn't that imply that PCS is not configuring systems correctly?

No, all PCS do is load windows to a OOB (Out Of Box) state, but the user still has to configure windows correctly.

My Recoil IV runs really hot - the keyboard is getting to around 45 degrees C when gaming and the area above the keyboard hits about 53 degrees C. I haven't checked the temperature of the bottom, but it's way too hot to have the laptop sitting on my lap - and it was configured by PCS

You need to monitor temperatures, it may mean the paste job isn't sufficient. It doesn't mean it's a problem with the chassis, just that the system isn't properly configured - either software wise or hardware wise. It's something that needs troubleshooting and correcting, if you just leave it that way it will only get worse and likely cause failure.

With any custom PC/laptop there is always going to be some configuration required by the user. It's inherent with custom builds. They're nothing like an off the shelf product.

 

[CraigBarrett215]

No, all PCS do is load windows to a OOB (Out Of Box) state, but the user still has to configure windows correctly.


You need to monitor temperatures, it may mean the paste job isn't sufficient. It doesn't mean it's a problem with the chassis, just that the system isn't properly configured - either software wise or hardware wise. It's something that needs troubleshooting and correcting, if you just leave it that way it will only get worse and likely cause failure.

With any custom PC/laptop there is always going to be some configuration required by the user. It's inherent with custom builds. They're nothing like an off the shelf product.

But then what constitutes correct configuration? I'm not aware of any Windows settings that impact how hot the machine runs. Though I've also never had a machine that runs hot enough that touching it is not the best idea. And how would I be able to tell if the paste job was done properly without taking the machine apart, then requiring a repaste anyway/

 

[SpyderTracks]

But then what constitutes correct configuration?

It’s just standard windows configuration, so make sure all updates are applied and all drivers are accounted for. On latest gen if windows isn’t configured correctly the system will run hot.

And how would I be able to tell if the paste job was done properly without taking the machine apart, then requiring a repaste anyway/

Standard thermals monitoring with hwmonitor or equivalent. Check testing temps (makins ure cpu usage is actually resting) for both cpu and gpu, then check load temps also.

 

[CraigBarrett215]

It’s just standard windows configuration, so make sure all updates are applied and all drivers are accounted for. On latest gen if windows isn’t configured correctly the system will run hot.

The problem is knowing what constitutes correct configuration. I've seen people on here talking about changing the max CPU power in power settings to 99%, but that makes no difference on my machine. So, not knowing which specific Windows settings affect temperatures, it'll be a lot of hours of guesswork in the hopes of finding something that actually works.

All drivers are the latest ones available from PCS, so if those are wrong then there's no fix from a drive point of view.

 

[SpyderTracks]

All drivers are the latest ones available from PCS, so if those are wrong then there's no fix from a drive point of view.

Did you manually install them then?

 

[CraigBarrett215]

I had tried reinstalling audio, video and chipset. But, as expected, they were all the versions that were already installed. Not really surprising since it'd be pretty weird to set the laptop up with drivers that are not the latest ones available.

Anything demanding (I played some Borderlands 3 on high settings) pushes the CPU and GPU temps up to 70+ (usually not getting above 73, though I've spotted them spiking up into the 80s very occasionally when BL3 was set to ultra). The main drive is sitting above 50 with the secondary usually 8 degrees or so lower. The area above the keyboard and the area of the keyboard around the < key get to over 50 and other parts of the keyboard range down to a little over 40 with most sitting close to the mid-40s. The different modes made almost no difference to the temperatures, nor did turning on the fan boost. In office mode the temperatures were about 1 degree lower.

When under minimal load (the CPU doesn't seem to go lower than 4%, GPU sitting at 0%) the CPU sits at about 45 and the GPU at about 40. The drives hover around 40 for the main drive and 32 for the secondary. The area above the keyboard and the warmest parts of the keyboard are at around 37 degrees, ranging down to mid to high 20s for the coolest parts.

These numbers are what I was seeing in Gaming Center. CPUID HW Monitor and Open Hardware Monitor give the same values.

 

[ubuysa]

I had tried reinstalling audio, video and chipset. But, as expected, they were all the versions that were already installed. Not really surprising since it'd be pretty weird to set the laptop up with drivers that are not the latest ones available.

Anything demanding (I played some Borderlands 3 on high settings) pushes the CPU and GPU temps up to 70+ (usually not getting above 73, though I've spotted them spiking up into the 80s very occasionally when BL3 was set to ultra). The main drive is sitting above 50 with the secondary usually 8 degrees or so lower. The area above the keyboard and the area of the keyboard around the < key get to over 50 and other parts of the keyboard range down to a little over 40 with most sitting close to the mid-40s. The different modes made almost no difference to the temperatures, nor did turning on the fan boost. In office mode the temperatures were about 1 degree lower.

When under minimal load (the CPU doesn't seem to go lower than 4%, GPU sitting at 0%) the CPU sits at about 45 and the GPU at about 40. The drives hover around 40 for the main drive and 32 for the secondary. The area above the keyboard and the warmest parts of the keyboard are at around 37 degrees, ranging down to mid to high 20s for the coolest parts.

These numbers are what I was seeing in Gaming Center. CPUID HW Monitor and Open Hardware Monitor give the same values.

The only way to know for sure whether your overheating is hardware or software related is to do a fully clean install of Windows from bootable media, deleting all existing system partitions (there are four of them), and allowing Windows Update to install all drivers, with the possible exception of the Nvidia driver. If you have a very new motherboard it would be wise to have the chipset driver handy in case Windows Update doesn't yet have it (which is unlikely). Device Manager will tell you whether Windows Update found all drivers.

Bearing in mind that any new Windows install runs hot - because there is a lot of post install configuration that goes on under the covers (disk indexing for example) - but generally only for the first 24 hours. See what your temps are like with just a clean Windows install, no other software, no other configuration, just vanilla Windows and drivers.

 

[CraigBarrett215]

I don't see how it's possible to test the temperatures with just a clean install of windows because not having the machine doing things other than initial setup doesn't put the machine under any pressure. The temperature when I initially turned the machine on was the same as the general temperature it sits at now when it's not doing much.

Unfortunately I don't see myself having time to experiment to figure out what specific driver is causing high temperatures or if it's just that case that the Recoil IV runs hot. It certainly wouldn't be the only gaming laptop to have uncomfortably high temperatures. There's so much time available to RMA the machine. And that a vendor is happy to send out machines that might not be correctly is actually a reason to return the machine, not to spend time working out what the vendor didn't bother to do.

 

[ubuysa]

I don't see how it's possible to test the temperatures with just a clean install of windows because not having the machine doing things other than initial setup doesn't put the machine under any pressure. The temperature when I initially turned the machine on was the same as the general temperature it sits at now when it's not doing much.

That's a fair point, but you want to limit what software you add in order to avoid reinstalling the problem.

Unfortunately I don't see myself having time to experiment to figure out what specific driver is causing high temperatures or if it's just that case that the Recoil IV runs hot. It certainly wouldn't be the only gaming laptop to have uncomfortably high temperatures. There's so much time available to RMA the machine.

Clearly RMA is an option you have and it's entirely your call.

And that a vendor is happy to send out machines that might not be correctly is actually a reason to return the machine, not to spend time working out what the vendor didn't bother to do.

That is flat out untrue and you should retract it. All builds are tested, if it had been exhibiting overheating symptoms at PCS it would not have been shipped. That's a fact.

 

[Paddy Baxter]

I have this laptop too and it is hot. I had the Defiance with a 980m and the proteus with regular 2070. Also both hot. 90 plus on cpu and 80-85 on the gpu. None of them were faulty. The fact that these PCS custom builds are often hundreds of pounds cheaper than big name brands is not just coincidence.

I called PCS about my hot 2070 super in the recoil that was hitting 87 degrees and thermal throttling. They said that under their tests the GPU never exceeded 80. By default, the gpu runs between 1500-1800 MHz when gaming according to msi afterburner, and this sent my GPU to its thermal limit in minutes, so I was confused as to how they could say it never exceeded 80?

However, looking up the Gpu on notebook check revealed that the upper range of the official boost clock was 1380 MHz. So I locked my gpu clock at 1380 MHz. This lowered my frames by around 5-10 percent depending on the game but the temperature (no coincidence by my reckoning) Topped out at 80 exactly. So this was probably more in the range of their testing, which makes sense that they would conduct their stress tests based solely on the manufacturers guidelines (in this case Nvidia)

 

[CraigBarrett215]

That is flat out untrue and you should retract it. All builds are tested, if it had been exhibiting overheating symptoms at PCS it would not have been shipped. That's a fact.

The configuration of the laptop was almost certainly not modified by the delivery company.

 

[ubuysa]

The configuration of the laptop was almost certainly not modified by the delivery company.

That's not an apology. The forum rules specifically prohibit defamation. I advise you to retract and apologise.

 

[CraigBarrett215]

You said that overheating indicates the laptop configuration may not be correct. But for the laptop to be incorrectly configured on arrival it'd have to have left the vendor incorrectly configured or have been modified after it left the vendor. It's clearly possible that, given the pressure PCS are under to get machines out to customers, there's a higher likelihood of errors occurring. So the number of new machines coming out that appear to be the subject of temperature complaints could simply be a function of the current situation. Obviously that's not the same as a vendor simply shipping machines that they know are not configured correctly.

The temperatures have not changed since I first started up the machine after receiving it - the temperatures under normal use are the same now as they were then. So it appears that no changes that have been made since I received it have had an impact on the temperatures at which the machine is running. The implication is that in testing the temperatures I'm observing are the same ones observed in testing, so they must have been considered acceptable.

 

[ubuysa]

You said that overheating indicates the laptop configuration may not be correct. But for the laptop to be incorrectly configured on arrival it'd have to have left the vendor incorrectly configured or have been modified after it left the vendor. It's clearly possible that, given the pressure PCS are under to get machines out to customers, there's a higher likelihood of errors occurring. So the number of new machines coming out that appear to be the subject of temperature complaints could simply be a function of the current situation. Obviously that's not the same as a vendor simply shipping machines that they know are not configured correctly.

The temperatures have not changed since I first started up the machine after receiving it - the temperatures under normal use are the same now as they were then. So it appears that no changes that have been made since I received it have had an impact on the temperatures at which the machine is running. The implication is that in testing the temperatures I'm observing are the same ones observed in testing, so they must have been considered acceptable.

Thank you for acknowledging that PCS are not 'happy to send out machines that are not configured correctly'.

 

[CraigBarrett215]

No, all PCS do is load windows to a OOB (Out Of Box) state, but the user still has to configure windows correctly.

So far there doesn't seem to be anything that reduces the temperatures.

I have this laptop too and it is hot. I had the Defiance with a 980m and the proteus with regular 2070. Also both hot. 90 plus on cpu and 80-85 on the gpu. None of them were faulty. The fact that these PCS custom builds are often hundreds of pounds cheaper than big name brands is not just coincidence.

I called PCS about my hot 2070 super in the recoil that was hitting 87 degrees and thermal throttling. They said that under their tests the GPU never exceeded 80. By default, the gpu runs between 1500-1800 MHz when gaming according to msi afterburner, and this sent my GPU to its thermal limit in minutes, so I was confused as to how they could say it never exceeded 80?

However, looking up the Gpu on notebook check revealed that the upper range of the official boost clock was 1380 MHz. So I locked my gpu clock at 1380 MHz. This lowered my frames by around 5-10 percent depending on the game but the temperature (no coincidence by my reckoning) Topped out at 80 exactly. So this was probably more in the range of their testing, which makes sense that they would conduct their stress tests based solely on the manufacturers guidelines (in this case Nvidia)

I'm currently trying to get Afterburner set up to show the stats while playing - I've been quickly going back and forth between the game and the desktop to check on temperatures on the CPU, GPU and SSDs. The temperatures are largely the same for all three modes with only office mode not pushing the GPU beyong the 1380MHz boost clock. And I've seen the same drop of around 10% in frame rates with that. I'm going to check again this evening with Borderlands 3 on ultra because that was the only thing that pushed the temperatures above that. But as far as I remember it pushed the temperatures above that regardless of the mode the laptop was set to.

How did you lock the clock to 1380MHz? I did some brief searching online, but couldn't find anything for it. Disabling overclocking in the BIOS will presumably do it, but is there another way?

 

[Paddy Baxter]

So far there doesn't seem to be anything that reduces the temperatures.



I'm currently trying to get Afterburner set up to show the stats while playing - I've been quickly going back and forth between the game and the desktop to check on temperatures on the CPU, GPU and SSDs. The temperatures are largely the same for all three modes with only office mode not pushing the GPU beyong the 1380MHz boost clock. And I've seen the same drop of around 10% in frame rates with that. I'm going to check again this evening with Borderlands 3 on ultra because that was the only thing that pushed the temperatures above that. But as far as I remember it pushed the temperatures above that regardless of the mode the laptop was set to.

How did you lock the clock to 1380MHz? I did some brief searching online, but couldn't find anything for it. Disabling overclocking in the BIOS will presumably do it, but is there another way?

once you get afterburner installed, hit Ctrl+F to access the custom clock/voltage curve. Here you can select one of the dots (of the lowest voltage possible) and slide it up or down to until it reads 1380 on the left of the graph. Then hit Ctrl+L to lock the clock/voltage at 1380. X out of the graph and hit the little ✅. I think my voltage is around 740 or something I’m not at home now so can’t check, but give that a go. The idea is to get as low a working voltage as stable at the 1380 mark. I think I even went as high as 1480 with temps maxed at 83

 

[CraigBarrett215]

once you get afterburner installed, hit Ctrl+F to access the custom clock/voltage curve. Here you can select one of the dots (of the lowest voltage possible) and slide it up or down to until it reads 1380 on the left of the graph. Then hit Ctrl+L to lock the clock/voltage at 1380. X out of the graph and hit the little ✅. I think my voltage is around 740 or something I’m not at home now so can’t check, but give that a go. The idea is to get as low a working voltage as stable at the 1380 mark. I think I even went as high as 1480 with temps maxed at 83

Great, thanks. I'll give it a try and see how the machine behaves.

 

[debiruman665]

I have a few points to make,

The out of the box overclock settings for the chips in a laptop are designed for the best possible benchmark results, temps be damned.

With a K series chip you will need to tweak the processor so that the boost clocks behave in a sensible manner. You might find that by reducing the max core multiplier down by a few hundred MHz will do wonders for temps and barely 0 difference to the frame rate.

I'm happy to assist OP in tuning the laptop if he wants. I spent months getting my own into a state where it would be silent as a mouse unless it was doing anything stressing.

 

[CraigBarrett215]

I have a few points to make,

The out of the box overclock settings for the chips in a laptop are designed for the best possible benchmark results, temps be damned.

With a K series chip you will need to tweak the processor so that the boost clocks behave in a sensible manner. You might find that by reducing the max core multiplier down by a few hundred MHz will do wonders for temps and barely 0 difference to the frame rate.

I'm happy to assist OP in tuning the laptop if he wants. I spent months getting my own into a state where it would be silent as a mouse unless it was doing anything stressing.

Any help will be much appreciated. I've tried locking the clock on the GPU to the spec'd boost of 1380MHz. That's reduced the max temperature under heavy load to 70. Which is pretty much the minimum under heavy load too unless I turn on fan boost in game and turbo mode in Gaming Centre. With the boosted fan running and the GPU locked the temperatures come down to a pretty calm range of 60 to 62 degrees.

Watching the temperatures under normal load, however, the CPU is still going up to 50+, which seems hotter than it ought to be. It comes down by several degrees if I set a custom fan speed profile for Office mode in Gaming Centre, but that of course makes the fans quite a bit louder.