The RIGHT way to test for stability

[nyker96]

thanks for sharing your research on stability testing. I have one point of contention, the occt linpack for my own experience fails much sooner than prime95 does and anything that passes small fft prime 95 8+hrs may not pass occt linpack. this is personal experience with 2 machines. yours may vary. still very good info.

 

[Diogenes2]

Originally posted by: error8

...... I had this dilemma for some time: should I use Linpack to test for stability, or should I just take the frequency OCCT is giving me as stable? But for some reason, I just can't stand or admit that my computer BSODS in Linpack,when I run the test. I can't get over that instability, created by a program that over stresses my cpu in a way that no other program on earth would stress it, in a normal use situation. It's simply not stable! I know that I could possibly take my Q6700 for example at a "stable" 3.7 ghz with a safe voltage applied, but it doesn't pass Linpack and that means that is just not absolutely 100% stable. I can't sleep with that thought in my mind. It's driving me mad. It has to pass Linpack. Curse you Intel for your stupid test. They should have kept it for themselves. Why was it offered to us? Why?

Is Linpack what you use your computer for ?

Is your computer a mission critical server ?

If your computer is stable while doing what you do, day in and day out, then it is stable..

On the other hand, what does overclocking get you ?

Would 3.5 or less, serve you just as well ?

 

[CoinOperatedBoy]

Originally posted by: ShawnD1
Another thing I think should be added to the top is run OCCT's power supply test. After running that, OCCT discovered that my power supply has incredibly bad voltage regulation when everything is running full blast. A bunch of individual tests might not catch something like that, and you're left wondering why the computer only crashes when trying to run a specific game.

It's also a nice way to test if your UPS can handle your overclocked system at full load.

This may be a good start, but it's no substitute for monitoring with a multimeter. The reports generated by OCCT were wildly inaccurate on the PSU test for me. I almost replaced my supply before I used a multimeter and found that my voltages were rock solid. Thanks to the Power Supplies forum for urging me to do that.

 

[ShawnD1]

Originally posted by: E4300
The use of Linpak or equivalent to "roast" the CPU does not simulate a real-world condition. Passing Linpak tells me that the CPU is okay at +100F ambient. It does not guarantee system stability with other applications under normal use.

We use Linpak because it's the hardest test any application could possibly have. There's no sense in stability testing with a Hello Kitty game then watching the computer blue screen when doing something more complicated.

 

[E4300]

Originally posted by: ShawnD1

Originally posted by: E4300
The use of Linpak or equivalent to "roast" the CPU does not simulate a real-world condition. Passing Linpak tells me that the CPU is okay at +100F ambient. It does not guarantee system stability with other applications under normal use.

We use Linpak because it's the hardest test any application could possibly have. There's no sense in stability testing with a Hello Kitty game then watching the computer blue screen when doing something more complicated.

Linpak's primary mode of failure is HEAT. Linpak does not guarantee system stability. Therefore, why over-volt the CPU to pass Linpak only to have issue with another application?

 

[ShawnD1]

Originally posted by: E4300
Linpak's primary mode of failure is HEAT. Linpak does not guarantee system stability. Therefore, why over-volt the CPU to pass Linpak only to have issue with another application?

Heat is the only reason our processors screw up in the first place. This is why we spend so much time and money on cooling, this is why sites like frostytech exist, and this is why you can overclock a processor to 6ghz if you cool it properly.
http://www.youtube.com/watch?v=wB0JodKgZ0A

You're doing a heat test because that's 99% of the battle.

 

[E4300]

Ahhh...but I'm not aware of any mainstream program that would place such a high thermal load on the CPU. Therefore, you're testing for a failure mode that will never happen in the real world. High voltage kills the CPU faster than high temperature!

The difference between a good and excellent air-cooled CPU cooler is about 50MHz of overclock. That's noise in my book.

 

[error8]

Originally posted by: Diogenes2

Originally posted by: error8

...... I had this dilemma for some time: should I use Linpack to test for stability, or should I just take the frequency OCCT is giving me as stable? But for some reason, I just can't stand or admit that my computer BSODS in Linpack,when I run the test. I can't get over that instability, created by a program that over stresses my cpu in a way that no other program on earth would stress it, in a normal use situation. It's simply not stable! I know that I could possibly take my Q6700 for example at a "stable" 3.7 ghz with a safe voltage applied, but it doesn't pass Linpack and that means that is just not absolutely 100% stable. I can't sleep with that thought in my mind. It's driving me mad. It has to pass Linpack. Curse you Intel for your stupid test. They should have kept it for themselves. Why was it offered to us? Why?

Is Linpack what you use your computer for ?

Is your computer a mission critical server ?

If your computer is stable while doing what you do, day in and day out, then it is stable..

On the other hand, what does overclocking get you ?

Would 3.5 or less, serve you just as well ?

Yeah, I know, I should just forget about Linpack, but like I've said, I just can't get over the part that it does fail at its test. It's a psychological thing.

And about frequency, the higher it is the better. Everything runs faster with the cpu clocked higher.

 

[elconejito]

What about stress testing tools for linux?

 

[Idontcare]

Originally posted by: ShawnD1

Originally posted by: E4300
Linpak's primary mode of failure is HEAT. Linpak does not guarantee system stability. Therefore, why over-volt the CPU to pass Linpak only to have issue with another application?

Heat is the only reason our processors screw up in the first place. This is why we spend so much time and money on cooling, this is why sites like frostytech exist, and this is why you can overclock a processor to 6ghz if you cool it properly.
http://www.youtube.com/watch?v=wB0JodKgZ0A

You're doing a heat test because that's 99% of the battle.

Heat is the "only" reason!? :shocked:

News to me. I could have sworn those isothermal shmoo plots had something to say about minimum stable voltage versus clockspeed at a given temperature.

Damn, all those years in college and industry experience were for naught, I should have just waited it out and boned up on my electrical engineering at the AT forums :laugh:

 

[ShawnD1]

Originally posted by: Idontcare
Heat is the "only" reason!? :shocked:

The voltage is 100% in your control so that should never be an issue. Look at your voltage vs clock speed graph, set the appropriate voltage and clock speed, heat test it with linpack. My processor could probably run at 4ghz if it was at 1.5v, but it fails the heat test and I turn down the voltage and clock until the heat test no longer fails.

 

[drizek]

The current version of OCCT does not have a ram testing mode(or at least I can't find it), just smal/medium/large data sets. Which is the one that is best for testing NB?

I agree with the others, this thread should be stickied.

 

[poohbear]

drizek to answer your question "large data set" is the best one.

this thread is over 2 years old an hasnt been updated? no mention of NB chipset testing for PhenomII's? (ie to overclock the NB/CPU controller which gives good performance increase for PhenomIIs)