Are stress tests necessary?

[tigersty1e]

Let's say I do a modest OC on my system...

2.8 GHZ (400*7)

Let's say it runs stable for hours and hours while I play games and it's super stable... is it possible to damage my equipment?

 

[MarcVenice]

If it runs to hot, it could decrease the lifespan of the CPU. I'd at least want to know how hot it runs. As for stability, it might not crash during gaming, but it might when you are doing something remotely important, which is something you want to prevent from happening.

 

[tigersty1e]

I'll def watch the temps and run stress tests for about 1 hour.... can't take too much time away from gaming.


I'll be done with some important work in about a couple of weeks and even if I do important work, I'll bring it back to stock.

But I read somewhere on these OCing forums that somebody said even though you run stable for days while playing games, you could be damaging the timing/circuitry on your system because the stress tests are designed to bring out these flaws. ...BS?

 

[MarcVenice]

I'd say BS, stresstests can't bring out flaws like things that would otherwise be damaging the timing/circuitry. You could be pushing to much vcore, and hurt the chip and perhaps the ram too if you push to much vram. Damaging the mobo itself would be really difficult. And your cpu will most likely overheat and throttle down before it dies because of to much vcore

 

[Saiyukimot]

I'd say the lower the overclock, the shorter stress test is needed.


I OC'd my Q6600 to 334x9 (3006MHz) on stock voltage with a Freezer 7 pro and after 4 hours 40mins with all 4 cores 100% on 2 orthos instances it was at 58degrees on coretemp.

Stable to me, and it didn't need another 6 hours or even 20 hours to tell me that.

 

[VirtualLarry]

If you care about a stable system at all, then stress tests are necessary. Even at stock clocks, it's a good idea to use them to burn in a new system. You never know when something might be out of whack.

 

[CTho9305]

Passing stress tests doesn't mean you're not damaging the system; failing stress tests doesn't mean you are. If you undervolt the CPU, it'll most likely wear out slower, but may fail stress tests (the circuits slow down, so a calculation might not finish within a clock cycle causing a test to fail). If you overvolt a system and keep at a reasonable temperature (and don't raise the clock speed), wear mechanisms like NBTI (negative bias temperature instability) and TDDB (time-dependent dielectric breakdown) will still be significantly accelerated, but it will pass any stress test you can throw at it.

 

[biostud]

not necessary but, a good indicator on whether your o/c is stable. It really depends on how much damage a system crash will do to you, the CPU will not be more or less damaged if it crashes during a stress test or while you're finalizing your most important report you're just about to finish.

 

[TC91]

tigersty1e, how much voltage do u have it running on? 2.8ghz to me doesnt seem to be a huge overclock on the 6300. as long as your temperatures are acceptable, i would think that the chip's lifetimie would indeed be decreased, but it probably will die after u decide u need a new chip. i guess it depends on how long u want to keep the chip.