Hardware Support, Please? [GA Inside]

Have u checked your PC for dust on the processor or elsewhere

Did you recently update your GPU drivers? I have a GT740 that constantly crashes if I install the latest Nvidia driver.

Try cleaning your drivers with DDU and then installing an older version to see if it makes a difference.

heat issue

Clean out the PC (get the dust out). Most common problem in these cases is overheating due to dust buildup or fan breakage.

Then run a memtest overnight, check for memory errors that might be exposed only when you truly push the machine... Maybe run a stress test and see if it reacts similarly if the memtest doesn't find anything. There's different ones for CPU and GPU, and some that stress both. Now if it crashes here it might seem like it would be the CPU or GPU that has problems - but more often it's actually the PSU - power supplies can degrade over time and give less stable power, or have caught one too many power fluctuations and now lets them straight through, or whatever - end result, not enough power delivered during high pressure.

So yeah, first de-dust, then check ram, then try with an alternative PSU if you can... That's about the order I'd go about it.

Start from monitoring your temperatures - running some game in windowed mode while having temperatures open in another one could tell you what is overheating and to what temperature.

Hard resets or permanent freezes can be caused by pretty much any component overheating, but it's GPU most of the time. It could also be CPU, PSU, Mobo or RAMs, in this order from likely to unlikely.

your PSU is?

When you say "Full power down", it means the computer power off? This could be a bad PSU. But these kind of trouble is hard to debug, a friend of mine had one similar issue but was the motherboard.. it´s hard to diagnostic without spare parts to test and replace.

Id be thinking power supply as well, its most likely on its last leg.

• Right click to "Start", open Windows PowerShell (admin), write "sfc /scannow"
• Use Ccleaner and clean your registry, remember to remove it afterwards because I've heard that it's not a safe software nowadays.
• Remove your GPU driver from Programs and Features, then use DDU (Display Driver Uninstaller) in safe mode and remove everything, restart your computer normally and install newest drivers, only install Graphics Drivers and PhysX because 3D drivers and Audio drivers are unnecessary and may even create problems with other drivers.
• Reset your bios to default settings, only change your SATA mode (IDE if only HDD, AHCI if you use SSD)
• Check startup programs, disable whatever you're not actively using (OneDrive, Acrobat Update etc.)

But before all that, check your temperatures. If a non-demanding indie game makes your CPU run 70+ degrees there're definitely an issue there. Clean dusts with canned air or with a compressor, change your thermal paste with something good (Arctic MX2-4).

If none of those solve your issue, use memtest86+ and leave it running overnight to see if there's something wrong with your RAM and CPU. I recently had a similar issue with my Phneom II 1075T x6 3.0GHz and it turned out to be dying so I upgraded.

if you have access to another psu, you could try changing them and playing a bit with the new one to see if it's that.
just don't go buy it blindly in case it's something else...

How many years with the PSU? How dusty It is? I´ve heard stories of people saying that the TR2 sometimes might just pop, just like somebody told you above (can´t confirm tho).
Have you changed your CPU thermal paste lately? or you´re still with the stock cooler from day one? how many years If this is the case?
Seasonic or Antec power supplies are reliable.

Hope you can find the culprit.

I'd also say it 's your PSU. However, you should also test a different GPU.
It could be that under certain circumstances it tries to draw too much power and then the PSU shuts down. Try a different GPU or your 660 in another PC.

Aside from the hardware suggestions here, I would also recommend looking at disabling any unneeded windows services and/or auto-start applications as well. And if you haven't done so already, disable Steam from downloading while you are in game (occasionally I enabled this setting and forget to turn it off until a resource hog runs into problems).

But since you specifically ask about hardware, from what is described above sounds to me like it could be the PSU or RAM. If you don't have spares lying around to try swapping out, then I find RAM to be easier to test/rule out first. Many Linux distros will allow you to boot from a disc or flash drive and many even include memory testing tools. I mention this because if you notice slow down outside of games at all, it may be worthwhile to compare if you encounter the same slowdown from another OS (all without touching your Windows install). I find Linux Mint is pretty easy if you're already used to Windows. If you aren't interested in using Linux to help diagnose, you can also try memtest x86+ as a standalone memory test (a word of warning though... it will run indefinitely until you shut it off; my recommendation is to just leave it running overnight and see if it has found any issues by the time morning rolls around). Another thing to try might be running some benchmarking software and see if any of those flag anything hardware-wise.

I will also ask about one thing I have not seen mentioned above:
What is the state of your hard drive? If you have an SSD, I recall hearing about issues if at least a certain percentage of the drive isn't kept free. The percentage seems to vary depending on who you ask, but generally I hear at least 20% of the drive should be kept free. If you have a traditional HDD, Western Digital and Seagate are reputable brands but even they are prone to failure eventually. I have seen older hard drives have issues with speed and especially if you hear a mechanical clicking sound it could be on its way out.