Computer hard freezes randomly - possible hardware issue

Question

Symptoms:

• When the computer is running, it randomly freezes and requires an hard reset. This happens both when using intensive applications (games, compiling) or just browsing the web.
• The freezing happens on both of my operating systems: Win 8.1 x64 and Arch Linux x64.
• One particular freeze on Win 8.1 happened while I was talking on TeamSpeak to a friend. After 1 minute I was able to continue speaking to my friend, even if the screen was frozen and nothing else seemed to work. The only thing I could do was literally speak to my friend through TeamSpeak. Alt-F4 wouldn't close TeamSpeak either. During the whole time the screen was frozen.
• After rebooting, sometimes nothing happens (the PC is on, though). I sometimes need to turn off and on the PC multiple times to get anything to display on the screen (the screen stays off otherwise, as if there is no input arriving to it).

What I have tried:

• Touching everything inside the case with the PC on. Tried touching RAM, GPU, SATA cables, etc. Nothing happened.
• Checking temperatures during stress/non-stress periods. They seem fine.
• Checking event logs/kernel logs. Nothing there.

Any idea what the problem could be? I suspect it is the GPU, because of the "TeamSpeak freeze". But honestly, I'm just shooting in the dark.

Info:

lspci -v, lshw

Solution 1

Since you have this problem on both OS, I'd recommend you to check all hardware step by step, especially motherboard. As for me, I would check it in following steps:

1. South Bridge temperature. You can just touch it carefully. If it is hot enough to burn a bit or close to it, that's bad sign. (Thermal sensors could lie, your burnt finger won't)
2. Memtest at least 4 passes. I had cases when 1st was OK and second or third shown errors.
3. MHDD just in case of bad blocks, that could freeze any system if situated near system core files. But that's unlikely your case anyway.

Both 2 and 3 steps preferably should be run from bootable media rather than from OS.

If nothing reveals faulty element, then just try to change components one at time (if you have them enough to test; buying just to test isn't good idea as it would likely cost as whole new PC). Also AIDA64 software has stress-test component, I was using it to find some issues and could recommend it too.

And I would add I had such a problem long ago. After long research I found faulty Ethernet network card driver. In my other case I had laptop beeping on boot about RAM fault and it was solved by replacing HDD. Keep in mind that sometimes problem is hiding in totally different component than you thinking of.

P.S. I'm sorry for a puzzle-headed answer, your question is too generic to provide exact instruction.

Solution 2

The main suspects are loose connections, faulty power supply, and thermal failure.

Loose connections are best diagnosed by correlating the failures with physical movements of the hardware.

Thermal failure will not always trip a temperature sensor. I had one machine which had undeniable thermal failure (including scorch marks on the CPU package) but whose temperature readings were never high. Apply some stress-testing such as memtest86+ (see the Arch wiki for more).

There's no easy way to test the power supply. If your supply is dodgy, you could benefit from using a cheap UPS.

Solution 3

I once had a Windows computer in which the display sometimes froze, but that everything else continued to work. However, sometimes I then received a message saying something like "Display driver restarted by system", and then everything returned to normal.

Some (very) few times I managed to get out of this state by entering Ctrl+Alt+Del and then Esc. Some (also very rare) times just leaving the computer alone for several minutes "fixed" it.

My theory at that time was a toss-up between bad video card or bad driver, or both.

You seem to have two display adapters : Nvidia GeForce GTX 275 and Intel Corporation 2nd Generation Integrated Graphics. My best advice is to update both your display drivers on both Windows and Arch Linux:

• For Nvidia on Windows, you should go to the NVIDIA Driver Downloads page and use the option "Automatically find drivers for my NVIDIA products" (requires Java or Internet Explorer).

• For Intel on Windows, see if this download page corresponds to your card.
  In any case, run Windows Update and see if it shows a newer driver in the optional updates section.

• For Nvidia on Linux, you might find a newer driver in the Unix Drivers page (or not).

• For Intel on Linux, see if Intel Graphics Drivers for Linux contains anything newer than your driver.

If updating to the latest drivers doesn't help, this might be a hardware fault. If your computer is still on warranty, use it to get a replacement video card (or motherboard for the on-board graphics).

Comments

[Vittorio Romeo]

memtest didn't trigger any issue. The problem still happens, and totally randomly.

[doz10us]

I was unable to post links to Southbridge definition and AIDA64 in answer itself since my reputation is too low. Here they are.

[mnmnc]

I think it's the RAM either way. If the Video Card would malfunction I think he would see some graphical artifacts like green screen, horizontal stripes etc.