Originally Posted by thlnk3r
Drin, I think this was discussed a couple of pages back? Shoot it might of been another thread but anyways...everyone has their own testing mythology. I prefer OCCT for 1hr on High and if that passes then I go onto Orthos (priority 9/blend) for 24hrs. Really truthfully nothing is 100% stable. It'll never happen, even at stock settings. That's the way I see things
Cheers, and yeah I was just being silly - I do realise the need for extended stability tests and how they can be useful, I just do not believe that someone should question what someoneâ€™s measure of stability is. If someone's system is not completely stable, or stable enough for their purposes, then it is their burden. There's potentially a lot of variables intricate to the actual instance of the chip itself, which may influence its ability to OC - claiming that it may be misleading to people hoping to buy the chip (jimbigood) should not be an issue. People should not buy chips expecting to 'overclock' something by a certain degree, because its never guaranteed due to the nature of overclocking itself. I know everyone does at some point, but they should always be aware of the 'risk' of overclocking expectations.
I'd estimate at least two thirds of people with an i7 claiming 4.0ghz never, ever offer any 'tangible' proof - its just their word. Reading the 955 thread, the poster who claimed 4.1ghz stable never offered any either, at least none that I saw, other than a cpu-z verification. In the end, you have to take everything with a grain of salt, and if someone posts greater evidence of their oc, massive props to them. You cannot however demand that of everyone, or belittle people who fail to meet your 'criteria' of stability since once again, it is all subjective.
Anyway, rant aside (no offence intended to any parties) I was primarily curious as to what people thought were good measures of long-term stability. Reading up on some posts done by Chew*, I figured that I'll start prime95'ing from now on. And to put my earlier statement of 'the risk of overclocking expectations' I was not able to hit 4ghz stable.
I do believe however that I did not mount my cooler properly, although the retention mechanism is very solid there are allowances for error; I have not had much experience with it yet and I believe its not seated optimally. The temps just do not make sense, I'm hardly seeing a decrease at all after:
- having my cpu lapped (flat, tested)
- installing 2x 120x38 fans in push/pull (to be fair I was using a single one on my noctua before my megahalems, but still)
- shrouding said fans
- moving my system to an open bench (tbh, now that my case is heavily modded I'd hardly see a difference but still)
- upgrading my cooler to what many argue is the best air cooler around
So no decent screenshots after all that effort
I did manage to hit 3920 stable (3 hours prime) which is a very slight increase in core clock, same voltages, but at the expense of my NB clock and RAM timings, so I figured its not worth posting over the one I posted a day ago. I'd spend more time reseating, benching, figuring out the bios, but with finals in a week I just have no time. Good news though, in three weeks I'll spend time to reach my final OC and leave it at that for good.
Best of luck everyone, and gl on your oc's!