Jenko32

It does not say you need to update in steps, you can go straight to the latest version.
The latest version will probably have more memory compatibility and it may be stable with EXPO 1 profile, so I'd try that first before anything else.

Right so I tried updating the firmware too, still happened. So I decided to try the laptop for an extended period of time and surprisingly it happened again - don’t know how I missed it the first time testing it, feeling a little stupid now. It must be the monitor then and I’m going to have to try send it back.
 
Thanks for the responses and tips though!

I increased the reocvery speed and it was able to copy pretty much all the readable parts in a few hours. Thanks

Nothing to say about the build but check that DP cable, it says it is VESA certified but it doesn't appear in the official database https://www.displayport.org/product-category/cables-adaptors/
Might be fine but would get another since it's going to be around those £ anyway. I have a KabelDirekt (which appears on the website) and I see it's on amazon.co.uk too
Also doesn't the monitor come with one? That should be fine if they give it to you with it

I would say to try to make the image before any physical step, so nothing more is lost if something goes wrong 

Fair enough, but headlight polish is pretty safe. It's a lot less abrasive than toothpaste and other homemade "scratched disc" remedies.
 
Attempt a rip as-is, then attempt physical fixes.

I would say to try to make the image before any physical step, so nothing more is lost if something goes wrong 

I would say to try to make the image before any physical step, so nothing more is lost if something goes wrong 

I finally completed one of my biggest bucket list items visit the LTT office. Now only if I met Linus that would be the highlight of my life. 

Assuming this is the original Dying Light, and not Dying Light 2, that is sort of odd. The 9700K should have more than enough resources to keep up with that game. And it looks like you aren't running out of RAM, so it's not an issue where the CPU is struggling to manage swapping data in-and-out of storage.
 
You say your PC is loud. Is your cooling working properly? If your 9700K is thermal throttling, that could cause it to be at 100% utilization here. What are the clockspeeds and thermals looking like?
 
If you don't have a program already installed to see CPU temps, I'd recommend HWInfo64.

Seems like nothing else other than Dying Light itself was using the PC, you should set up MSI Afterburner to see temperature and fan RPM of CPU and GPU (if you have a 3 fan gpu it may have 2 different speeds)
 
Adding: Overheating usually is a recurring problem, other than keeping in mind that higher framerate that can cause the GPU temperature to rise, you should check if any fan is stuck or dead

Not all PCIe pins are used in GPUs, some manufacturers try to save money by not placing the contact to be gold plated. What card is it? Try to find some pictures online if you fear you might have damaged some of used ones
For example this is an older GPU that didn't use all the pins https://m.media-amazon.com/images/I/71eXqVo3bfL.jpg

Not all PCIe pins are used in GPUs, some manufacturers try to save money by not placing the contact to be gold plated. What card is it? Try to find some pictures online if you fear you might have damaged some of used ones
For example this is an older GPU that didn't use all the pins https://m.media-amazon.com/images/I/71eXqVo3bfL.jpg

Hello Fellow Cooling Enthusiasts!
 
I conducted an  experiment recently and I wanted to share the results.
Hopefully someone here will find them helpful or interesting. 
 
 
Earlier this year, I had to buy a replacement graphics card for my 'work from home PC".  I enjoy moding, water-cooling, and experimenting with computers.
I wasn't able to get a graphics card that had any thermal-related bottlenecks, so I didn't expect to get any performance boosts right now.  
But if my concept worked, I could transfer it to a better graphics card later. And in the meantime, I could reduce fan noise.
 
 
I needed something 'relatively easy' and 'relatively cheap' and didn't want to mess with or replace the existing heatsink or fans.
I bought a used EVGA GeForce RTX 3060 XC GAMING, 12G-P5-3657-KR.
At the time, no one made a waterblock for this 170 watt card, which is understandable.  Bykski makes one now, but it is $189.99!
 
 
So this got me thinking, is there another simple way to supplement the existing GPU cooling system?  
 
I started by using the EVGA X1 utility to max out the power and temp targets.
 

 
My Armor 9 phone has a great FLIR camera, so I loaded up Prime95 and just watched the graphics card heat up.  
Ambient temperature 73F and I did all my testing with the case panel removed, so my case airflow was below normal.
 
 
My GPU has an aluminum backplate which was almost entirely at ambient temperature except a little hot spot smaller than the GPU processor, but centered at the same location.  
 
I took off the back plate and saw a small thermal pad touching 4 little components (see photo down below).
 
This pad was about 1mm thick.  So the backplate was already being used as a heatsink, but only for those 4 small components.
 
Then I wondered:
1) How hot does the PCB get directly underneath the GPU processor?
2) How well does that heat transfer through the PCB board?
3) Could I transfer more heat to the aluminum backplate?
4) Could I then effectively remove the heat from the backplate?
 
I found a few forum posts that suggested transferring heat out the backplate was possible, but they all added more heatsinks and extra fans which I didn't want to do.    
 
So here is what I did instead:

I drilled and countersunk two small holes in the aluminum backplate.
 

 
I found an old $5 VRM water-block (bought on clearance) and drilled two matching holes in it.  
 
I tapped the holes in the VRM waterblock, then sanded the copper side with 600 and 1000 grit sandpaper on a sheet of glass.  
(I wasn't trying for a mirror finish, just removing some milling marks in the copper.)
 
Then I screwed the waterblock to the backplate after applying a thin layer of decent thermal paste.
 
I added 4 large thermal pads (13W/mK, 2.5mm thick) around the existing thermal pad, then reattached the backplate to the graphics card.
 

 
 

 
I reran Prime95 (without any liquid in the waterblock) to see how much heat was getting through the PCB, through the thermal pads, and into the backplate.
 
Before adding the thermal pads, the aluminum backplate was almost entirely at ambient except that small hot spot next to the processor. 
After adding the 4 extra thermal pads, without changing fan settings, the entire backplate was at least 30F above ambient with the hottest zones 46F above ambient.
 

 
 
So it appears that a decent amount of heat can be transferred through the PCB, through the thermal pads, and into the backplate.
 
Next I connected the VRM waterblock to my custom water loop and restarted Prime95 and let it run until temperatures stabilized again.  
 
Now the entire backplate was much cooler, especially closest to the processor chip.
 

 
 
I was able to retune the graphic card fans, and now they hardly ever spin up past 30%, even under load.  
 
I ran some GPU benchmarks before and after this mod, but the improvements were small, as expected with an RTX 3060.
However, I suspect that if this mod were applied to a higher wattage card, it could potentially help quite a bit.
 
I think the experiment was a great success!
 
The PCB may not be an efficient thermal conductor, but it does get very hot.  Large thermal pads were able to conduct the heat into the aluminum backplate.  Some of that heat can be dissipated away from the backplate with good PC case airflow, but a small waterblock works much better.
 
Supplementing the cooling of a graphics card through the backplate is 'relatively easy', 'relatively cheap', and can move enough heat to be worth doing.

 
 

I would suggest to reinstall drivers using DDU and reinstall Valorant so you can exclude all problems except Windows. If you still have the problem and it's not a problem with an update from Valorant itself then the problem is likely Windows or a program you recently installed
 
Does it happen with other games?

I fixed the issue.
My motherboard has 6 ram slots, i was using 8gb x3 in wrong configuration.
so whenever my ram usage was going above 16GB it restarted my computer. 
played around with the ram slots, got the right ones and now even with ram usage 20+Gb i never had any single crash. 

Lancool II Tempered Glass eATX Full Tower Computer Case - Black
SuperNOVA 850 GA 850 Watt 80 Plus Gold ATX Fully Modular Power Supply
B550 MAG Tomahawk AMD AM4 ATX Motherboard
Ryzen 7 3700X Matisse 3.6GHz 8-Core AM4 Boxed Processor with Wraith Prism Cooler
Trident Z RGB 32GB (2 x 16GB) DDR4-3200 PC4-25600 CL16 Dual Channel Desktop Memory Kit F4-3200C16D-32GTZR - Black
Radeon RX 5700 XT Red Devil Overclocked Triple-Fan 8GB GDDR6 PCIe 4.0 Graphics Card
Platinum 2TB SSD 3D QLC NAND M.2 2280 PCIe NVMe 3.0 x4 M.2 Internal Solid State Drive
BarraCuda 8TB 5400RPM SATA III 6Gb/s 3.5" OEM Internal Hard Drive
GIM KB-24 RGB Case Fans, 6 Pack 120mm Quiet Computer Cooling PC Fans, Music Rhythm 5V ARGB Addressable Motherboard SYNC/RC Controller, Colorful Cooler Speed Adjustable with Fan Control Hub
S107 Gaming Keyboard and Mouse Combo w / Mousepad
Monitor is a Viotek GNV34DBE 34" Curved Gaming Monitor
 

Soldering wires is beyond me. I guess I'll have to use some led strips near my fans to get the same effect

Wonder why Nvidia or Amd haven't tried to build a function like this into their cards. h264 is built into Nvidia cards for decoding and encoding stuff for Shadowplay. Why not use it for encoding an incoming video signal?

During the benchmark, do you see a stutter every few seconds or it's just low fps?
The system has the same drivers as the old GTX 1060 or other cards? Try DDU to exclude drivers problem

When it happened they just arrived home and were entering the building so they got away in time.
Look these videos if you want to hear more about what happened
 

Update:
I got a single stick of DDR4 RAM from Amazon, popped it in to see if RAM was the issue. System has the same symptoms.
 
I'll see if I can get my hands on some AM4 CPU to see if it's the issue... which it basically has to be, right? The only way it could be anything else (to my knowledge) is if I got a motherboard that stopped after 7 months, and one that wasn't working 100% correctly out of the box, which seems like low odds.
 
I don't know what I'll do if I try the CPU and that doesn't fix it... so I hope it will. Thanks Jenko32 for your help.

I don't think we'll know accurately what happened as an autopsy is not doable. From what he said the battery main switch, which switch off if the current draw is more than 63A, was just turned off since they just arrived to destination and the battery pack just started shooting flames
The pack was specifically made for him by an Aliexpress vendor, as he didn't find anything premade for what he needed. Both battery pack were made for 80A but he pulled only peak 40-50A (reminding that there's a switch that limited to 63A)
He did the math and there was about 2.4A per battery, and they should be rated for 7A continuous

Theoretically the PCI Express delivers max 75W
The sata cable should be rated for max 54W
The 1650 has no need for an additional PCI Express Power, it uses only 75W
1650 Super uses 100W, so 75W from PCI Express plus 25W from the sata cable could be ok
The RX570 uses max 150W, some says 135W, so let's use 135W for this scenario 
75W+54W=129W so mathematically it could be fine as it's only 5W missing, but it's not actually.
Every cable as a specified thickness that is going to allow to pass the amount of the needed current (Voltage X Current=Wattage). If you don't use the correct thickness of cable you could be overheat the cable, following melting of the outside rubbery sheath, it catching on fire or it running at 100° or even higher, lowering its lifespan and possibly causing other problems
The cables from the sata cables, even though of being capable of passing 54W do not have enough cable thickness to avoid the things I mentioned above.
So use it at your own risk with a 150W card but should be fine with a lower one

In my HTPC I use the integreted GPU of an i5 4670, the HD 4400, previously I used to have a Q9500 paired with a GeForce GT210 and both have no problems with Netflix, Prime Video and 4k H265 videos
Edit:Actually the Q9500/GT210 combo got laggy in 4k h265
Basically get anything that was made after 2016