E8400 Overclocked to 4Ghz - Is this alright?

Exactly +1

Most obsured thing I have ever heard of in my life. Almost as obsured as someone else thinking its okay to shoot 1.75vc through a 45nm Chip.

 

Pointless in which sense? Are you suggesting you think you will catch any errors that may occur by only running for 30 mins?? That is simply not the case, I'm afraid.

Alternatively, if you're saying you are willing to have a less stable system than you would have at stock clocks, that is of course your perogative but you are making a choice to comproise potential stability for performance.

 

No, I'm saying that with a test like IBT, if you have it selected to max ram and leave it for 20 - 25 runs.....there's very little chance of it erroring after that. And even then all it will prove is that your CPU can successfully run Linpack.

Don't get me wrong, I'm all for a stable system...I use mine for Cosmosworks and starccm+ CFD so it needs to be stable. I just think some people take the stability testing thing way too far and give it too much credit, because as I said, it only ensures your CPU can run IBT fine. For example I've had a system 14 hours prime stable only to crash in 2 minutes after playing CoD4.

 

Yep, no processor is the same, mine is stable at 4Ghz with 1.18v

 

hmm, it seems like people are saying they don't *think* there's any point running beyond 30-60mins because they believe it's unlikely to error. This really is not good test methodoloy

Personally, I very much make use of 30-60 mins runs to home in on potentially stable OC configs but then do an overnight run to test fully (and, over the years on different rigs, I most certainly have gotten fails at this point).

 

Over the years on different rigs you probably didn't use Linpack though....it's a different beast altogether. The entire point of Linpack is that it doesn't take a day to do it.

 

Good OC, im still peeved i need 1.55v just for 3.8Ghz ¬_¬

 

so i keep hearing people mention linpack. i assume its another stress test like prime95. which should i run?

 

You have a PM.

 

I never said that I used it for 24/7 clocks. I ran around 1.35v for 4.5ghz for 24/7. However, I know a couple guys who have terrible c0 chips and run 1.5v though them for 24/7 and have had no problems at all using that amount of voltage.

 

The beauty of the linpack based stress tests such as the Intel Burn Test and the LinX is the short test duration. The linpack can adequately test the CPU stability in 5-20 minutes as compared to 8+ hours required for the Prime. It can save a lot of time when tweaking. Not good for stressing the chipset and the memory though. Overnight Prime large fft (2048-4095) non in-place is the best for chipset/memory testing.

 

I agree I use IBT for on the fly testing and prime blend 11hrs+ to test the chipset and ram, also memtest for overclocking ram

 

alright thanks so much for the help guys. Right now I'm at 4229.8mhz with 1.352V. Just started Prime 95 with small FTT. I'm post results once it's done(going for 6 hours at least).

 

lol yeah i was wondering for a min :nerd:

 

guys I think I may have done something very very stupid. I was running Prime95 for about an hour or two and got impatient. I decided to attempt 4.5ghz. it booted up but the resolution was like at the lowest, and my PC wouldnt detect my graphics card. I freaked out and went back to bios to set my fsb back for for 4.2ghz.

I rebooted and everything seems fine. resolutions is back to normal too.

 

You should be okay. Funny things like this can happen when boot into windows with a very unstable overclock.

 

just means it was unstable. You can't really break anything just by changing CPU frequency alone. Worst that can happen is that it will not POST...in which case you'll have to reset the CMOS. Big deal.