Video DXGKRNL_FATAL_ERROR

[NavBK]

My laptop crashes into a BSOD randomly, even when idle, with the Video DXGKRNL Fatal Error. It does not occur when I disable my NVIDIA GTX1650 graphics card. The steps I have tried - 

• sfc -scannow and dism restore health
• Running DDU and performing a fresh reinstall of the NVIDIA drivers
• Complete factory reset
• Turning off ASPM for PCIE
• Turning off Hardware Accelerated Scheduling

This issue has been going on for some time now, and I have been trying to solve this on and off, resorting to just disabling the card when I have no choice. My local service store was clueless regarding this issue, and a friend of mine with the same laptop also face the same issue, whose service person just disabled the graphics card as a duct tape fix. Please suggest a solution. 

[BeemerBiker]

Either the GTX1650 chip is overheating or it has gone bad.

How to use air duster from WIKI how
https://www.wikihow.com/Clean-a-Laptop-with-Compressed-Air

General instructions for cleaning from HP tech.
https://www.hp.com/us-en/shop/tech-takes/how-to-clean-your-laptop#

This HP video shows how to clean air vents on your laptop.

You might be able to check for GPU/CPU temperature problems using this tool (GPU-Z) and selecting 'sensors'

There is also a Hot Spot that can exceed the 100c temperature limit. This automatically throttles the GPU

if overheating you may need to renew the heat sink compound on the GPU or replace the fans or both.

[NavBK]

Thank you for the response. I did another complete reset of the system, to double check that the issue is not any driver update. I used the gpu-z software to log the temperature, and it is around 50C (Hotspot around 60C). I usually run my laptop with a cooling pad. 
Also, I used to have NVIDIA statistics HUD active when the system used to crash while idle, and I just noticed in gpu-z that having it active is causing the GPU and Memory Clock readings to have values, compared to not having the HUD active, when there are no values. 
I ran a game to note the temperatures under load, and the values were GPU - 79C, Hotspot - 89C. Also, the CPU temperature hit 99C. But there was BSOD in this run.
After the game, while I had the HUD active, and I was running normal tasks not requiring the GPU, it crashed. GPU-Z logs does not show any variations from the idle temperatures mentioned above. The reliability history log is attached below.

This error used to occur along other than the dgxkrnl error when the issue initially started, but after that, it was only the dgxkrnl for a long time.

Then after sometime, it crashed again (again, not using GPU, did not even have the HUD active this time), this time with a new error: