Having trouble getting my i7 930 stable @ 4 ghz

[Ikthus]

I'd definitely do an overnight run of P95 blend to make sure your OC is completely stable. When I was attempting to stablize my 4GHz OC, I almost always got through 20 passes of LinX with no problem only to have 1 or 2 cores fail or get a rounding error a few hours into Prime. But yeah you shouldn't rely on just one program to test for stability, just take it from me...I had to reinstall W7 last week because my 4.2GHz OC which I thought was stable ended up corrupting my OS lol.

[rjbarker]

Just to clarify guys....I found LinX fine for stress testing 920 (Quad Core)...its the 980X Hexa Core it doesn't stress properly. hence using P95 for 6 Core .

[bvsbutthd101]

I'm try lowing some voltages to lower the temps

[nzaneb]

Stability is a relative term. If you're running a file server, workstation, folding, etc... sure you should prove stable with P95. But if you only edit a few pictures, watch movies, and surf the net (at 4ghz + speeds)... 20 passes of LinX is more than enough. Everyone's definition of stability is different, and no one's is indefinitely right. Just my .02$

[Ikthus]

Originally Posted by nzaneb

Stability is a relative term. If you're running a file server, workstation, folding, etc... sure you should prove stable with P95. But if you only edit a few pictures, watch movies, and surf the net (at 4ghz + speeds)... 20 passes of LinX is more than enough. Everyone's definition of stability is different, and no one's is indefinitely right. Just my .02$

I don't think anyone OCs their CPU to 4GHz just for web surfing or movies! I wouldn't consider just running 20 passes of LinX enough to prove stability though. No harm in spending a few extra hours with OCCT or Prime just for safe measure.

[nzaneb]

I think you may have missed my point. Example: All my benching rig does is run benchmarks. It simply has to be stable enough to finish a bench and get a screenshot. While my work station receives a few hours of P95 Blend. Why should they be subjected to the same definition of stability? Stability is measured by the intended use of the computer, not a Universal "stability" standard. If you're worried about loosing irreplaceable information if you suffer a crash... you probably shouldn't be overclocking to begin with. The best test is usually using the computer for what it is intended for (Gaming, Folding, Work, etc...)

[Chris]

I always thought oc is all about bragging right. To be on the same base, some define stability as P95 for 10 hours or more.

Why do I oc? Because I want to run my cpu at a higher speed, without which I will have to buy a higher rated cpu at higher cost. Sure, my photo post-processing can run at rated speed nicely (being a quad), but it is satisfying to know that it runs a few seconds faster. Besides, half the fun of building my own pc is to oc.