3 Months Later I'm Getting Errors

[wrathloki]

So 3 months ago or so I successfully overclocked my e6400 core 2 duo from 2.13 ghz to 3.2 ghz and I ran fractal tests for over 24 hours with absolutely no errors. Now for some reason today I noticed games erroring and freezing up so I tried doing some fractals again, and I get an error about 20 minutes in. I haven't changed anything since so what could be causing this? Is my chip failing?

I'm running:
-Gigabyte GA-965P-DS3
-Intel E6400 2.13ghz(3.2 oc'd) core 2 duo
-2 gigs corsair ram running at 1:1 synchronous mode
-factory overclocked 700mhz/2000mhz 8800gt
-seagate 500 gb hdd with 32 meg cache

 

[lucuis]

Perhaps. I'll throw some theories out there.

First thing that comes to mind is heat. 3 Months ago the general temperature was cooler.

Side Note: Unless your Voltage is crazy high i doubt it the chip failing. But ESD has been known to cause a slow breakdown in components that have been victim to it. So if you weren't very careful with grounding and static etc while building your computer that might be it.

Also PSU if they are gonna fail they tend to fail early on, download a program like Hardware Monitor or Everest to check your PSU voltages. For better accuracy haul out a multimeter. Again check your temps, compare them to your previous data from when you first OC your computer.

Let me know if any of this bring up something conclusive.

 

[Verillion]

I would recommend downloading Orthos. This program will let you stress the ram, stress the cpu, or do a blend test of both. Then you can correctly identify which of the two is the culprit.

Also, getting a program to monitor voltage while you are running these tests would be great. If you see a drop of more than .125 on your Vcore, then the PSU may be the culprit. There is a "feature" known as VDROOP, which in theory is a good thing, because it allows minor fluctuations to occur in the processor voltage, as to protect it from dirty power. However, if your power supply is not very constant the VDROOP might actually be a bad thing, causing the fluctuations to be magnified. I have seen that a .250 Vcore drop causing serious stability issues.

 

[Evilonigiri]

Orthos or Prime95, which ever you like. Run Prime95 blend test first for 24hours, then the small FFT for 24hours. If blend fails, it's normally ram, or perhaps mobo. If you're running stock setting for ram, then run memtest86 to check to see if your ram is defective or not.

Another possibility is that one or all of the following is failing: PSU, motherboard, and perhaps CPU. If it's PSU, the easiest way is to grab another one that works and use it. If the errors go away, its the PSU (duh).

 

[Shadow703793]

1st Run Memtest86+ for ~6hrs. If it's a RAM problem Memtest86+ will fail.

2nd, download Prime95, run small FFT for at least 8hrs. imo, 24hrs is an overkill.

 

[shadowthor]

Well prime95 can fail between the 8hrs and 24, so 24 hrs would be ideal to ensure stable operation.

 

[Mondoman]

While memtest86+ is a good basic test, it doesn't stress the CPU or memory like Orthos/PRIME95/etc. We need more info from you for educated opinions:
1) What is the VID of your e6400? What CPU voltage have you set in the BIOS ? What is the idle CPU voltage reported in the BIOS?
2) What is/was the temp of your CPU at idle and while running Orthos or similar dual-core stress program in "blend" mode?
3) What is the brand and model number of your power supply? Power supplies slowly lose capacity over time as they age; poorly designed/made ones may also show poor regulation even from the beginning.
4) One or more of your DIMMs might be defective. Make sure you have set the DIMM voltage in the BIOS to the value specified by Corsair for your specific DIMM model number. As shad suggested, run memtest86+ O/N and see if you get any errors. If so, remove all DIMMs but one and re-run the test through at least 3 complete passes. Repeat this for each DIMM to see if there are errors in only one DIMM -- if so, it's time to take advantage of the lifetime warranty on the DIMM.