Crash/freeze whenever I load/load into a game.

FORMAT

It can be windows "natural" self-degradation

 

Please RTFP.

 

think 10.1 or 10.2 was good and stable for many, for chipset, maybe u can use windows drivers, sry been awhile since i had ati

 

I changed to 10.2 it is more stable but I still get the same error. Now running Geek3D FurMark (works now!)

It almost direct went up to 70c and is hanging arround there now let's see what happens next.

Edit : I ran it for 1 hour, gave 100k frames, I will do a longer test tommorrow I will go sleep now.

 

APB uses unreal 3 engine, maybe u could test another game with same engine
this might pin problem down to driver or settings, if no unreal3 runs on ur pc
(cant entirely rule out hw error tho)

List_of_Unreal_Engine_games#Unreal_Engine_3

(atleast America's Army 3.0 should be free)

other 2 games uses some odd engine, cant say much about those

 

For some reason I didn't have problem last night, or at least scarcely. Now I have the problem all the time again, I have no cleu where to look next... CPU/Memory seems to be fine.

If noone is able to see a different problem then I will buy a new GPU.

 

Check idle-load voltages yet, just for quietness in future.

 

17-120

What I now have gives

61-189

Also the dB on the full load on the card I saw (MSI HD 6870 Twin Frozr II ) is only 31

 

I mentioned not Your PC quietness, but quietness in our further discussions .
Perhaps Your PSU can't to operate normally. You can use HWMonitor http://www.cpuid.com/downloads/hwmonitor/1.18-setup.exe. You will able to see Min Max voltage values.

 

I also suspected the PSU, because the crashes/freezes he had that forced him to hard reset the system wasn't a BSOD, but a black screen.

He already tried Prime95, which failed after a few hours (which shouldn't be; Prime95 isn't that rigorous compared to IntelBurnTest).

 

I had a BSOD when I crashed first time, after 5½ hours, I will redo the test now, and post the results later on.

Test has already been running without any heat/pc problems for 2 hours.

 

With Prime95, if your system is really stable, nothing should be wrong for around 8+ hours; a more thorough test usually lasts a day.

With IBT, if you survive 10 runs (without differing 'Residual(norm)' values), it'd usually mean your system would survive more than 2 days of Prime95.
If you want to try IBT, try it first without OC-ing anything; I don't want to be hold responsible if your system goes up in smoke.
No, seriously, that's just how harsh IBT is; make sure to at least do a Standard test with it.

 

It has run for 8 hours now, without any problems.I will try IBT tommorrow to make sure.

 

also suspect psu, only way to make sure is to test with different one

usb sounds can be dodgy too

 

My USB worked perfectly fine the last month, untill 1½ week ago, and I can't change my PSU, all the other I have are to low Watt to help.

 

u can combine 2 psus, power on other thats not connected to mobo with wire trick

 

I tried to run it, it crashed stopped in literally a second it says to following :

Conclusion : your Pc has faile dlinpack testing.

... it is undeniable to conclude that your CPU and/or RAM is unstable.

I tried to run the test in full RAM usage, just wondering if my CPU is stable with the Prime95 test I did, does this mean my RAM is faulty?

 

If CPU / RAM is being said unstable, even at default clocks, means that one of them (or both) isn't getting enough power.

So, here's what you've done so far.

• With Prime95, even when it's possible to survive an 8 hours run, has failed before on just around 5 hours.
  This means that your CPU and/or your RAM is fine, but something else is making it unstable.
• Memtest run went fine, so definitely nothing wrong with your RAM.
  But this test doesn't take much power out of your system.
• Furmark fails, even after switching drivers.
  The only difference is how long you can run Furmark before it fails completely.
  This means something is making your VGA unstable.
• IBT fails right after starting the test.
  Something is really wrong with your system, and it's definitely not your RAM that's at fault (it survived Memtest, remember?).

Actually, there is another test suite that also uses Linpack : OCCT.

If you're unsure whether your RAM is at fault or not, then you can run Memtest once more.
If the new Memtest run doesn't fail, then I think you should change your PSU.

 

Memtest didn't fail I just did a test it ended up at 103.5% coverage without a single problem.

I just ran OCCT is said my DirectX was outdated, I updated it to the newest and ran the linpack test. After 1 minutes of checking it ran up to high usage and in almost 40 seconds it told my CPU is turning to hot.

The result (here) is pretty staggering, it almost inmediatly tells me I have problem with my heat while other programs fail to mention it, even Prime95 didn't have problems.

Funny is that all the voltage don't really change apart from the 12V and the VCore. All the processor have same results so I just added 1 picture of them.

Just wondering what my next step should be, replacing my CPU changing fan or run more tests?