Running 'CPU Burn-in' & 'SETI@Home' at same time to test out O'C- is this OVERKILL

[thatsright]

Okay, now I'm trying to find a good FSB clock rate that is stable to O'C with my ABIT IC7 & P4c 2.4Ghz & Kingston HyperX PC3500. Just finished reading an article in MaximumPC that talks about Stress Testing utilities for O'Cing. They recommended using CPU Burn-in, and a few others. Well I thought why not run CPU Burn-in AND SETI@Home at the same time for about an hour or so. I would do this to see if my BIOS settings are a good candidate for a long term (24 Hours) Burn in peroid. But does anyone think that even doing this would give me 'false' crashes when running both programs at the same time? I was running both, as just running one Stress Test would only give me 50% CPU utilization, so I wanted 100 CPU% usage.

Well I ran both for a while AND Juno Email client, & then I opened up one IE6 window and my PC just re-boot on its own. Should I just run one stress test and that?s it, or try to run two simultaneously?

Thoughts, Opinions, Suggestions, etc.

 

[Slammy1]

I have different tiers to stress testing. I run my games on the first tier (the joys of o/c'ing), and do a second tier which is similar to what you describe. I keep adding programs until my CPU usage gets over 90%. I try to run a lot of multimedia along with Prime95. For example, my current o/c (220FSB, actually 2-3-2-5-4, I have HyperX on a P4P800) ran about an hour with Asus Probe (Probe errors are a common o/c problem), control panel, WinAmp playing mp3s with video, WMP playing a Divx file, ATI TV, and WinDVD (I watch the DVD to look for errors). The third tier is Prime95, which you can run 2 instances of to move CPU utilization to 100% (which I never do), for longer periods (ca 12 hours). So, yes, I'd say your current config has some instabilities .

 

[thatsright]

Okay just for the Record: This is what my settings are and what I've been running. 2.4C @ 276Mhz FSB= FSB of 1105Mhz. CPU speed of 3314.7Mhz. RAM is Kingston HyperX PC3500 512MB X2 running @ 2-3-3-7 @ 2.7Vdimm, RAM speed is 221Mhz.

As for the stress testing, my goal is/was to run the CPU at 100% usage for whatever stability testing I will be doing. Be it 2 hours or 24. Right now I have the CPU Vcore at stock of 1.525, so I'm not really worried about frying it. The hottest the CPU has gotten was 64 C during a two hour run of SETI@Home and CPU Burn-in.

But is running the overclocked CPU at 100% for like 12 or 24 hours just asking for a dead CPU?

 

[WarCon]

No, that is how I tested thermal compounds. I used CPUBurn (never seen anything give temps like it) combined with Prime95 (to make sure I was stable still). The Prime95 actually cooled the processor by about 1C from what CPUBurn does alone. But if you have a weak processor CPUBurn will finish it.

I started CPUBurn and then started CPU Burn-in. Then I stopped CPUBurn and watched the temp drop by 2-3C. So your probably safe with CPU Burn-in.

Here is the homepage, but use at your own risk as the page says........ http://users.ev1.net/~redelm/

 

[Noid]

I first run memtest86 overnight to make sure I have no memory issues.

Then ,,,, I run Prime95 , a hardware monitor, and 3dMark2001se on looped demo mode with sound on for 24 hours.

I do this because P95 wont catch 3d performance errors. (overheating and weak voltages for the mb)

click my 220dual linky

 

[WarCon]

Originally posted by: Noid
I first run memtest86 overnight to make sure I have no memory issues.

Then ,,,, I run Prime95 , a hardware monitor, and 3dMark2001se on looped demo mode with sound on for 24 hours.

I do this because P95 wont catch 3d performance errors. (overheating and weak voltages)

Agreed.....

 

[thatsright]

Originally posted by: Noid
I first run memtest86 overnight to make sure I have no memory issues.

Then ,,,, I run Prime95 , a hardware monitor, and 3dMark2001se on looped demo mode with sound on for 24 hours.

I do this because P95 wont catch 3d performance errors. (overheating and weak voltages for the mb)

click my 220dual linky

Thanks for the heads up Noid: I just tried to run 3dMark03 for like the 4th time or so with the FSB @280 and it just stops, and I get back to the Desktop. Now once I find a good idea of a FSB to test at for the long haul, I'll run Memtest-86 overnight.

BUT did you run Prime95+HardwareMonitor+3dMark2001se(looped demo) ALL 3 at the same time? If so, how long did you have all 3 go for, how many hours.

Thanks

 

[Noid]

You really need to run memtest86 FIRST.

=Your crashing to desktop could be caused by data corruption (memory - improper timings , more voltage needed, or just maxed out)

I typically run the P95 3dmark test 24 hours.
(dont forget to turn off screensaver and power saving.)

Afterall,,,, I hate problems,,, BDOD's, crashing, or freezing applications wont cut the mustard with me.