Computer problems

[frankdur]

Hello internet, I am having an problem with my computer where it crashes randomly. At any given time, my screens (2 moniters) will go black and my graphics card fan will go full speed. After 3 seconds of that my screen displays an no input sign and my computer crashes. All my drivers are updated. I am thinking its a graphics card problem, any ideas?
Specs:
Operating System: Windows 7 Ultimate 64-bit
System Manufacturer: OEM
System Model: OEM
BIOS: Phoenix - AwardBIOS v6.00PG
Processor: Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz (8 CPUs), ~2.7GHz
Memory: 6144MB RAM
Card name: ATI Radeon HD 4870 X2
Display Memory: 3825 MB
Dedicated Memory: 1014 MB
Shared Memory: 2811 MB
Current Mode: 1920 x 1200 (32 bit) (59Hz)
Monitor Model: Acer P243W

 

[MindoverMaster]

Try the monitor on a different system or try a different monitor on your posted system.

 

[frankdur]

My display works fine on a different computer, and I hooked up a different display to my computer and it worked, then again this problem occurs randomly.

 

[Slaymate]

Run FurMark for a few minutes and tell us what your temperature is peaking at.

 

[NiGHTS]

If the fan only starts to spin fast once the displays become disconnected, then I do not think its an overheating issue, otherwise the fans would be the first early indicators of a impending failure.

When you say "crash", I assume you mean having an abrupt reboot. Given your symptoms I would subscribe that your failure order starts with the display output completely failing while your GFX CPU fan spins full blast, at which point either your computer has already rebooted at the first point symptom of failure, or it does indeed wait 3 or so seconds before the reboot.

Does your GFX CPU fan rev up whenever you turn on your computer? If so, then its simply possible that your computer simply rebooted by itself, causing all those symptoms by nature of the POST of the graphics card.

Now if its unusual in normal events to hear that fan go full force, thats a whole other story. If this is the case, there is a chance that the temperature detection circuitry is faulty, causing bad readings. I have seen this happen on at least one computer I worked on, but it was on the motherboard instead. Beyond that, shorts or other on-board faults of the GFX card could cause the GFX CPU to issue a reboot command by way of the internal watchdog timer (in other words, its waiting for a response from something that is taking way too long and shouldn't be) causing it to issue a reboot on the GFX CPU, and then would cause the driver to become desyncronized with the mapping of the peripheral port, leading to the MOBO CPU to issue a fault which, depending on the severity of the drivers condition, leads to a full abrupt reboot instead of the usual memory dump procedure of windows.

There can be other things wrong, but if you find the fan spinning up unusual, even on computer power-up, then something seems fundamentally wrong with a circuitry. Heat would not do this (and if it was, you would notice it!), and if it was a loose heat sink with the GPX CPU, then the crashes would happen nearly at the same time after boot-up, rather than randomly. If you think it might be a loose heat sink or not enough heat paste, if possible check that out on the graphics card just to be sure that you ruled out that possibility.

Ultimately you need to do a few more tests to figure out what the true cause of the problem is. If you had another graphics card to test with that would be super helpful.

 

[frankdur]

ran furmark, got 68 C

 

[frankdur]

When you say "crash", I assume you mean having an abrupt reboot. Given your symptoms I would subscribe that your failure order starts with the display output completely failing while your GFX CPU fan spins full blast, at which point either your computer has already rebooted at the first point symptom of failure, or it does indeed wait 3 or so seconds before the reboot.

Does your GFX CPU fan rev up whenever you turn on your computer? If so, then its simply possible that your computer simply rebooted by itself, causing all those symptoms by nature of the POST of the graphics card.

The computer doesn't reboot, it just sits in this state of "nothing" the computer is still running but is doing nothing.

 

[NiGHTS]

The computer doesn't reboot, it just sits in this state of "nothing" the computer is still running but is doing nothing.

During the lock up, when you press "Num Lock" key, does it toggle the corresponding light on the keyboard?

Also, when you quickly press the power button during the lockup, does the computer turn off immediately or does it just sit there?

 

[frankdur]

hitting the numlock key does nothing and to shut off the computer i have to hold down the button (force shut down)

 

[NiGHTS]

hitting the numlock key does nothing and to shut off the computer i have to hold down the button (force shut down)

Sounds like the CPU is halted.

Does the fan ever rev up to full speed when the computer turns on?

If "Yes", the halt could have been issued before the video stopped working. And that means that the video card is probably not the initiator of the crash. At that point you could try looking for memory dump files or use the windows event log to try to find the driver that reported any kind of issues right before the crash occurred.

If "No", the graphics card itself could be at fault at a hardware level. A good follow-up test would then be to do the following:

1) If you have another available port for your graphics card, use that.
2) If you have any video acceleration features turned on in Windows for your graphics driver, turn that off. This also goes for DirectX settings or driver settings which could be stressing the card. Basically try to setup your card to be as simple as possible.
3) Go to the BIOS and dumb down your acceleration settings, including system RAM. Look for any video features that are available and turn anything you don't NEED off.
4) If your graphics card uses an additional power source from your PSU, use another compatible cable if available. The power feed itself could be rippled or under/over powered, which, while I highly doubt would cause the system to behave that way with its internal voltage regulators, could still be a problem. So if possible use another source of power. If you know how to use a multimeter, you could try to see if the power is accurate and steady from that cable.
5) If you are comfortable unscrewing the heat sink from the graphics card, It wouldn't be a bad idea to check if the heat transfer paste is on the GFX CPU needs replacing.

If all this doesn't at least improve the situation, try using another card entirely for a while at least to rule out the possibility of the video card.