GPU doesn't exist (driver problem? hardware problem?) and a sensor problem or something else?

[evlesoa]

I was told from SysNative to come over here and ask about what's happening with the sensors because Martin (that's the name they gave me in the thread) can tell me more about this.

Please refer to these two posts and the screenshots in there (they both have HWiNFO64 screens in there):

Constant system crash related to nvidia? Windows 10 BSOD

[...] When you return the system, include a printed copy of this thread to show them what crash experts have found in their system. Regards. . . jcgriff2 Definitely agree, especially since the system is only 2 months old as well.

www.sysnative.com

And

Constant system crash related to nvidia? Windows 10 BSOD

[...] When you return the system, include a printed copy of this thread to show them what crash experts have found in their system. Regards. . . jcgriff2 Definitely agree, especially since the system is only 2 months old as well.

www.sysnative.com

Thanks very much in advance.

 

[Martin]

Sorry, but I cannot view attachments (screenshots) on the linked forum without registering there.
Please post them here or upload to some public site, i.e. imgur.com

 

[evlesoa]

Sorry, but I cannot view attachments (screenshots) on the linked forum without registering there.
Please post them here or upload to some public site, i.e. imgur.com

No problem, I shall post them here.

This is what's happening now:



















Previously, sensors would turn on/off. The problem we're trying to figure out is if there's a hardware issue, an incorrect/incomplete Windows install, or a driver issue. The PC itself has been used for only two months, although I've officially had it for about three or four. I didn't turn it on until end of July, however. I'm trying to get an RMA from the company, but if there's a possibility that this is a driver issue and I can fix it, then no purpose to RMA it. There are a lot of issues so idk what's going on *really*.

When the graphics card *did* work, the sensors would flash on/off constantly. Now, as you can see, the problem has progressed into simply not reading anything. This happened when I tried to shut off the PC in the middle of a nvlddmkm.sys crash instead of letting it have a BSOD-like crash that it normally goes through.

 

[evlesoa]

Also, in idle with just the browser running (it does this even at the BIOS screen), this is happening. Sometimes, if I use the "wrong profile" (i.e., performance instead of entertainment), the fans just stay at 2k RPMs. Sometimes, the fans are completely shut off on the GPU side. And other times, if the fans do stay at 2k RPM, they only start spooling up when hitting 90+ c.



Seeing how HWiNFO is reporting basically nothing at all on the GPU, maybe you have an idea? Could it really be as simple as a sensor issue as the main cause of all of this? Can that be fixed without RMA'ing it? I use a cooling pad that runs 24/7 under the laptop. I also shut the laptop down at night. It appears to be an "unknown hardware problem", but could it be a driver problem alone? Maybe you have some idea. Thanks a lot!

Btw, CPU is always running @ turbo clocks with 1% usage. (this is with browser closed).







Sorry for so many screenshots and stuff, it's just that I don't know what's normal anymore for this laptop. I was told to lower my temps, but I don't see how I can. The idle has always been 50-60c.

 

[Martin]

Well, this is not easy to diagnose precisely as such analysis would require dedicated hardware diag tools.
But according to those symptoms (communication with the GPU seems to be failing), everything points to a problem with the NVIDIA GPU. This can be either a driver or hardware fault. Considering that you tried a clean install, different driver versions and the way how those failures appear, I believe this is a hardware failure.
That system is only 2 months old, there's nothing to hesitate - RMA.

 

[evlesoa]

Well, this is not easy to diagnose precisely as such analysis would require dedicated hardware diag tools.
But according to those symptoms (communication with the GPU seems to be failing), everything points to a problem with the NVIDIA GPU. This can be either a driver or hardware fault. Considering that you tried a clean install, different driver versions and the way how those failures appear, I believe this is a hardware failure.
That system is only 2 months old, there's nothing to hesitate - RMA.

I see. Very well. I'm waiting on them to green light the process after having sent them return address info and all other stuff. Once the ordeal is over, I can post back when I get a response from them as to what went wrong. I'm dying for a confirmation that this wasn't a driver issue.

 

[VictorVG]

evlesoa

According to your description, the picture is similar to the failure of the driver startup parameters. Try to remove it using the DDU (Display Driver Uninstaller, author forum page and official download link. No install needed, just unpak use 7-Zip and use . Notes: then unpack just skip "Display Driver Uninstaller.pdb" file - is debug data base for WinDbg debugger and don't needed for use DDU) utility, and after restarting the OS, install the driver again. I think if the reason is precisely the damage to its settings, then the kernel panic should go away. Well, otherwise you have to look who sits on what DOS interrupt - these conflicts affect the launch of the OS and can lead to its instability. I have come across this phenomenon more than once and the only solution here is selection of boards in the system board slots.

 

[evlesoa]

evlesoa

According to your description, the picture is similar to the failure of the driver startup parameters. Try to remove it using the DDU (Display Driver Uninstaller, author forum page and official download link. No install needed, just unpak use 7-Zip and use . Notes: then unpack just skip "Display Driver Uninstaller.pdb" file - is debug data base for WinDbg debugger and don't needed for use DDU) utility, and after restarting the OS, install the driver again. I think if the reason is precisely the damage to its settings, then the kernel panic should go away. Well, otherwise you have to look who sits on what DOS interrupt - these conflicts affect the launch of the OS and can lead to its instability. I have come across this phenomenon more than once and the only solution here is selection of boards in the system board slots.

I tried DDUing it many times. Running Unigine's Valley in DX9 mode crashes it in seconds, but on DX11 mode it crashes after a few minutes of running. I just get a TDR 141 error and that a "hardware error has occurred". Then the whole system attempts to shut down, but never goes into a BSOD. Instead, it turns off and then it says that the system "was unexpectedly shutdown". I tried DDUing, I tried reinstalling Windows in a variety of manners (e.g., keep settings, in-place upgrade, wipe everything, etc). Tried even using drivers from back in January of this year. Issue still occurs. As of right now, HWiNFO is completely blank on the GPU unless the GPU is utilized by something. That's new. I can DDU the drivers again, but I'm positive that it won't fix the issue and that the crashing will continue... I really tried so many different drivers, I started blaming myself for updating stuff on the computer >_< -- but it turns out that it doesn't matter anymore about what is updated... that's not the cause, I guess.

I'm still waiting for an RMA number. I'll give them a week before I start pestering them. lol.

 

[VictorVG]

Your laptop CPU is overheating - pay attention to the temperature sensor CPU Packedge == +98 (I marked it on your screenshot) - it means that the heatsink of the CPU cooling system is clogged with dust (in laptops it is usually carried out on the heat pipes outside the package of boards ), or heat is not removed from it, for example, thermal contact with the heat sink is broken, and under these conditions, microcircuits may fail due to overheating.

This is the first, second - you have a laptop, and Generic drivers may not be suitable for them. the manufacturer could modify the VLSI enable circuit or the firmware of the chips at its discretion, and then the Generic drivers from the GPU developer may work. Take the drivers from the download section specifically for your model from the website of the laptop manufacturer, remove the Generic drivers through DDU and put the ones you took from the site.

At the same time, it is not at all necessary that the drivers from the section for your country or the main section (the country where the company’s headquarters are located) immediately fit. It is possible, for example, that they will have to be selected for the Samsung P30, and the manufacturer’s technical support will not be able to help you solve the problem. This was exactly the case with this machine - the drivers from the AMD site and the Samsung Russia site for the ATI Radeon 4650 (rename on this notebook as ATI Radeon 5470 [SAMSUNG]) did not fit this APU (micro-assembly from the CPU and the GPU Radeon GPU in the common case), which caused one of the CPU cores to freeze and, as a result, the kernel would panic with emergency restart of the OS, and drivers with Samsung USA came up. Technical support and service centers knew about the problem, but could not find its cause, and only after my technical report were the drivers replaced on all servers with the necessary ones.

 

[evlesoa]

Your laptop CPU is overheating - pay attention to the temperature sensor CPU Packedge == +98 (I marked it on your screenshot) - it means that the heatsink of the CPU cooling system is clogged with dust (in laptops it is usually carried out on the heat pipes outside the package of boards ), or heat is not removed from it, for example, thermal contact with the heat sink is broken, and under these conditions, microcircuits may fail due to overheating.

This is the first, second - you have a laptop, and Generic drivers may not be suitable for them. the manufacturer could modify the VLSI enable circuit or the firmware of the chips at its discretion, and then the Generic drivers from the GPU developer may work. Take the drivers from the download section specifically for your model from the website of the laptop manufacturer, remove the Generic drivers through DDU and put the ones you took from the site.

At the same time, it is not at all necessary that the drivers from the section for your country or the main section (the country where the company’s headquarters are located) immediately fit. It is possible, for example, that they will have to be selected for the Samsung P30, and the manufacturer’s technical support will not be able to help you solve the problem. This was exactly the case with this machine - the drivers from the AMD site and the Samsung Russia site for the ATI Radeon 4650 (rename on this notebook as ATI Radeon 5470 [SAMSUNG]) did not fit this APU (micro-assembly from the CPU and the GPU Radeon GPU in the common case), which caused one of the CPU cores to freeze and, as a result, the kernel would panic with emergency restart of the OS, and drivers with Samsung USA came up. Technical support and service centers knew about the problem, but could not find its cause, and only after my technical report were the drivers replaced on all servers with the necessary ones.

It's definitely not clogged with dust. I can tell you that with certainty because I personally opened it up. Just how much dust can get into a 2-month old laptop? Enough to screw up the cooling capabilities? And regarding drivers, Clevo/ProStar/Sager do not release drivers specific to any country. They have a website that has a bunch of drivers for this particular model and that's it. They're a "one unit for the rest of the world" type of company.

Also, even in BIOS the fans cannot keep up with the max frequency of the CPU. The CPU is constantly throttled, surely. The fans are always loud in the BIOS. But that doesn't explain why I can cause instant crashes in Unigine's Valley or in any other benchmark or game from boot (immediately after). I can cause crashes at will in some other 3d software by loading the right thing. To give you an example of what my crashes are (since I no longer run 3D programs, I do not get crashes anymore, haha):



So, the page faults are a result of...? Bad drivers? But I've already tried using manufacturer-provided drivers, which is why I moved away from them. It doesn't change the crashing, it seems. I installed intel management engine on top of the drivers and some issues have changed from instant crash in a 3d app to a delayed one. Coincidence? Or a direct result? I can show you if you don't believe me. Just let me know and I'll DDU the drivers and then show you the TDR 141 errors that keep recurring as well in another post. I have to roll back to a driver from several months ago (that was the last update), or to drivers from almost a year ago from the mfg. From Day 1 of ownership, this laptop the Nvidia Inspector reported no information on the graphics card. The screen flickers, also. I don't know why. That was from day 1, too, but it seems installing updates makes it less worse. I don't know, maybe a perception issue. My last computer was an Asus laptop from 2013. It had an i5 3230m, 8 gigs of ram, 512gb HDD, and a GT 635m. I had no problems with Nvidia Inspector or any sensors reporting anything--there weren't any blanks compared to this newer laptop, and everything was stable. I used it all the way up to two months ago--when I switched to this because the keyboard was failing and it wasn't worth keeping at that point.

You probably want to see what the crash dump analysis was for my memdumps:

From the Stop 0x50 dump:

2: kd> k
# Child-SP RetAddr Call Site
00 ffff8108`39d29de8 fffff803`0e5e3463 nt!KeBugCheckEx
01 ffff8108`39d29df0 fffff803`0e4730bf nt!MiSystemFault+0x1d6733
02 ffff8108`39d29ef0 fffff803`0e5cf120 nt!MmAccessFault+0x34f
03 ffff8108`39d2a090 fffff803`285afc6f nt!KiPageFault+0x360 << We crash here!
04 ffff8108`39d2a220 ffffbe8f`77e79b78 nvlddmkm+0x6efc6f << Illegal page fault caused here
05 ffff8108`39d2a228 00000000`00001c43 0xffffbe8f`77e79b78
06 ffff8108`39d2a230 ffff6b28`c2e3db69 0x1c43
07 ffff8108`39d2a238 ffffbe8f`77e6c000 0xffff6b28`c2e3db69
08 ffff8108`39d2a240 00000000`00000007 0xffffbe8f`77e6c000
09 ffff8108`39d2a248 fffff803`2887429d 0x7
0a ffff8108`39d2a250 00000000`00000003 nvlddmkm+0x9b429d
0b ffff8108`39d2a258 00000000`00000000 0x3

2: kd> !pte ffffe50071f3f000
VA ffffe50071f3f000
PXE at FFFFA9D4EA753E50 PPE at FFFFA9D4EA7CA008 PDE at FFFFA9D4F9401C78 PTE at FFFFA9F28038F9F8
contains 0A000000067C0863 contains 0A000000067C1863 contains 1A000001437F8863 contains 0000000000000000
pfn 67c0 ---DA--KWEV pfn 67c1 ---DA--KWEV pfn 1437f8 ---DA--KWEV not valid

Since there is no corresponding physical page mapped, then we cause a page fault, however, the address provided is either completely invalid or within a non-pageable section of memory.

2: kd> lmvm nvlddmkm
Browse full module list
start end module name
fffff803`27ec0000 fffff803`29479000 nvlddmkm T (no symbols)
Loaded symbol image file: nvlddmkm.sys
Image path: \SystemRoot\System32\DriverStore\FileRepository\nvcvi.inf_amd64_11bdd2121036771e\nvlddmkm.sys
Image name: nvlddmkm.sys
Browse all global symbols functions data
Timestamp: Thu Sep 5 20:44:22 2019 (5D716596)
CheckSum: 0155B6AC
ImageSize: 015B9000
Translations: 0000.04b0 0000.04e4 0409.04b0 0409.04e4
Information from resource tables:

Your graphics card driver appears to be very new, I would certainly consider rolling back a few versions to see if it provides any system stability.

There is also a Stop 0x113, which is directly related to the Windows video sub-system crashing, unfortunately the bugcheck parameters aren't well documented at all.

Here's the call stack from it:

8: kd> knL
# Child-SP RetAddr Call Site
00 fffff485`f5dea2d8 fffff803`79ac41a0 nt!KeBugCheckEx
01 fffff485`f5dea2e0 fffff803`799e32a7 watchdog!WdLogEvent5_WdCriticalError+0xe0
02 fffff485`f5dea320 fffff803`798ba519 dxgkrnl!DpiFdoHandleSurpriseRemoval+0x167 << Likely couldn't handle a PnP IRP for some reason?
03 fffff485`f5dea360 fffff803`7981617a dxgkrnl!DpiFdoDispatchPnp+0xb9
04 fffff485`f5dea410 fffff803`80f99991 dxgkrnl!DpiDispatchPnp+0xea
05 fffff485`f5dea530 ffffa105`a9ce5000 nvlddmkm+0xd9991
06 fffff485`f5dea538 00000000`00000000 0xffffa105`a9ce5000

The first frame just saves the processor context and raises an error which causes the system to produce the dump file.

8: kd> .frame /r 1
01 fffff485`f5dea2e0 fffff803`799e32a7 watchdog!WdLogEvent5_WdCriticalError+0xe0
rax=0000000000001f11 rbx=0000000000000001 rcx=0000000000000113
rdx=0000000000000019 rsi=0000000000000000 rdi=ffffa105a11f3070
rip=fffff80379ac41a0 rsp=fffff485f5dea2e0 rbp=ffffa10596edd9a0
r8=0000000000000002 r9=00000000000010de r10=fffff80363bc10a0
r11=fffff485f5dea2b0 r12=ffffa105a8db8030 r13=ffffa10596930001
r14=00000000c00000bb r15=ffffa105a8db8030
iopl=0 nv up ei ng nz na pe nc
cs=0010 ss=0018 ds=002b es=002b fs=0053 gs=002b efl=00040282

8: kd> u
watchdog!WdLogEvent5_WdCriticalError+0xe0:
fffff803`79ac41a0 cc int 3 << breakpoint trap?
fffff803`79ac41a1 cc int 3
fffff803`79ac41a2 cc int 3
fffff803`79ac41a3 cc int 3
fffff803`79ac41a4 cc int 3
fffff803`79ac41a5 cc int 3
fffff803`79ac41a6 cc int 3
fffff803`79ac41a7 cc int 3

However, looking at the call stack again, I'm starting to lean towards either memory issues or a driver issue. The surprise removal call refers to the following IRP - IRP_MN_SURPRISE_REMOVAL - Windows drivers

 

[VictorVG]

evlesoa

Digital technology either works in the case, or basically does not work. There is no third. I showed you specific moments of CPU overheating in your screenshot, but you thought that it doesn’t matter. Well, wait for what happens next. With drivers is similar.

 

[evlesoa]

evlesoa

Digital technology either works in the case, or basically does not work. There is no third. I showed you specific moments of CPU overheating in your screenshot, but you thought that it doesn’t matter. Well, wait for what happens next. With drivers is similar.

I understand, but there's not much I can do about the power limiting. Nothing. I can only "rev it down" by using a different power plan. That's what I did. There's also a fan issue where the fans don't seem to be "very responsive". I asked for a BIOS update but I was turned down for one.

What alternative do I have? I'm just sending the unit back to have it fixed. This is beyond any driver update at this point as I understand it.

And revisiting that, I know my CPU is over-heating. A fresh windows install should *not* cause a CPU to over-heat like that. But what can I do about this? I was told that if a complete system reinstall with a media creation disk doesn't solve the problem, this is beyond fixing on the software level. This is why the pros @ SysNative told me that I need to RMA it. Nothing about how this machine is behaving should happen in the first couple months of usage. Show me a new machine that, with a fresh install of windows, misbehaves like this. This was the reply I got from one of the admins:

I would pester them to death about this system. From reading your posts, I gather that the trouble did not just start recently, rather it likely started day 1 after you initially booted up (or problems began at that time that ultimately led to the NVIDIA BSODs).

NVIDIA writes some of the finest drivers in the world; we rarely see them in BSODs as compared to ATI video drivers, which are definitely problematic. My estimate of the ratio for BSODs caused by NVIDIA drivers vs. those caused by ATI drivers - 1:100 - meaning that we may find one NVIDIA driver issue (almost always out-dated drivers) vs. 100 ATI driver issues.

The fact that your system is BSOD'ing and naming NVIDIA tells me that there is a hardware problem afoot. I do not believe at all that there are any problems with the current NVIDIA drivers. Your system's problem is more than likely unknown hardware failure.

If you paid by credit card, contact the credit card company and try to have the charge stopped/reversed. Doing so will certainly get AVADirect's attention and force them to deal with you.

You deserve a new system, considering that you likely began to have problems with this system very early on after receiving it.

Just to review - 4 BSODs - Bugchecks -

BugCheck 1C7, {0, ffffb60be5ba0000, a8a65024, 144da0}
Probably caused by : ntkrnlmp.exe ( nt!SmHpBufferProtectEx+10cb04 )
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``
BugCheck 113, {19, 2, 10de, 1f11}
Probably caused by : dxgkrnl.sys ( dxgkrnl!DpiFdoHandleSurpriseRemoval+167 )
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``
BugCheck 50, {ffffcc0150939000, 2, fffff80421d7fc6f, 0}
Probably caused by : nvlddmkm.sys ( nvlddmkm+6efc6f )
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``
BugCheck 50, {ffffe50071f3f000, 2, fffff803285afc6f, 0}
Probably caused by : nvlddmkm.sys ( nvlddmkm+6efc6f )
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``

0x1c7 - store component detected a corruption in its data structures; probable cause "NT" - the Windows Kernel - a default of sorts
0x113 - Microsoft DirectX graphics kernel subsystem has detected a violation
0x50 - invalid memory referenced; illegal page fault

When you return the system, include a printed copy of this thread to show them what crash experts have found in their system.

Regards. . .

jcgriff2

Would you say that you disagree with him? He carries badges and has the title of:
"Administrator,
General Manager, BSOD Kernel Dump Expert, Microsoft MVP"

 

[VictorVG]

evlesoa

You are in vain afraid of giving the car for warranty repairs and picking drivers for it. The fact is that any developer of an electronic circuit has the right to finalize a typical circuit for switching on a microcircuit published in the technical documentation in accordance with the tasks it solves. It is enough to put a chip instead of the one indicated on the typical wiring diagram similar in properties, but issuing other codes to the status registers so that the Generic drivers cannot work correctly with this device.

Therefore, I can assume that the video card in your machine is assembled according to the standard electrical circuit that has been improved relative to the NVIDIA partner documentation published, which led to the need to use a modified firmware for it (Video BIOS), and as a result to incompatibility with Generic NVIDIA drivers manifested in observables you phenomena.

Regarding CPU overheating, it does not require long-term operation for its occurrence, but a random violation of the thermal contact “CPU cover - radiator sole” is sufficient when assembling the machine, for example, as a result of not tightly pressing the heat sink, or errors when calculating the cooling system, for example, using a unified cooling system for the production of a family of models with different CPU's. The reason can be any, the main consequences.

But, in any case, overheating of the CPU in such a new product (in terms of life), if the user complies with the requirements for operating conditions specified in the specification, is the basis not only for contacting the manufacturer’s service center with the requirement for warranty repair, but also the legal basis for filing against the manufacturer of a lawsuit for the release and sale of a knowingly non-conforming approved specification of the final product, which qualifies as a factory defect.

And I’m sure that the manufacturer will prefer to quietly eliminate the marriage, and its representatives will catch you in the toilet with a bunch of apologies if only the matter would not end with a public scandal. After all, such a lawsuit against the manufacturer is not a desirable public scandal for him that instantly negatively affects not only his reputation, but also his profits, and shareholders in this situation will not forgive the board of directors for the drop in their income.

 

[evlesoa]

evlesoa

You are in vain afraid of giving the car for warranty repairs and picking drivers for it. The fact is that any developer of an electronic circuit has the right to finalize a typical circuit for switching on a microcircuit published in the technical documentation in accordance with the tasks it solves. It is enough to put a chip instead of the one indicated on the typical wiring diagram similar in properties, but issuing other codes to the status registers so that the Generic drivers cannot work correctly with this device.

Therefore, I can assume that the video card in your machine is assembled according to the standard electrical circuit that has been improved relative to the NVIDIA partner documentation published, which led to the need to use a modified firmware for it (Video BIOS), and as a result to incompatibility with Generic NVIDIA drivers manifested in observables you phenomena.

Regarding CPU overheating, it does not require long-term operation for its occurrence, but a random violation of the thermal contact “CPU cover - radiator sole” is sufficient when assembling the machine, for example, as a result of not tightly pressing the heat sink, or errors when calculating the cooling system, for example, using a unified cooling system for the production of a family of models with different CPU's. The reason can be any, the main consequences.

But, in any case, overheating of the CPU in such a new product (in terms of life), if the user complies with the requirements for operating conditions specified in the specification, is the basis not only for contacting the manufacturer’s service center with the requirement for warranty repair, but also the legal basis for filing against the manufacturer of a lawsuit for the release and sale of a knowingly non-conforming approved specification of the final product, which qualifies as a factory defect.

And I’m sure that the manufacturer will prefer to quietly eliminate the marriage, and its representatives will catch you in the toilet with a bunch of apologies if only the matter would not end with a public scandal. After all, such a lawsuit against the manufacturer is not a desirable public scandal for him that instantly negatively affects not only his reputation, but also his profits, and shareholders in this situation will not forgive the board of directors for the drop in their income.

Very interesting. It's that serious? I don't know how this stuff works. I'm genuinely surprised to have a physical hardware problem. I will say that that because I've never had a computer experience a hardware issue in the ~20 years of playing with computers.

So, what do you suggest I do on the graphics card part? I'm still waiting for an RMA number. It'll take a week or so, I'm guessing. I guess just DDU the drivers, right, and roll back to the mfg provided ones? I can try it. Do I have to do it in Safe Mode? Or can it be done in the regular booted environment?

Edit: I rolled back the drivers to manufacturer ones by going through specifically safe mode. Same issue with sensors.





HWiNFO. Take a look at the GPU on the right side:

 

[evlesoa]

Just wanted to give an update:
It was a motherboard failure of some sort. I had sent it in for an RMA because of what everyone told me. Tests run for several days by the company that sold the laptop to me would have the system haphazardly shut down. Not sure what part or anything was causing that (SSD, GPU, CPU, RAM, etc, were all tested). Because of other un-foreseen circumstances, it took over two months to get it fixed and sent back to me.

Now we know the problem, at least. Mobo got replaced and everything tested, and it works stably.

 

[Martin]

Thanks for your feedback !