highoc

Leave your commemts

__________________
[/CENTER]

TwoCables

Hax.

j/k Nice work, man!

__________________

highoc

thanks man thinking of going further

still on stock volts

__________________
[/CENTER]

TwoCables

Now I can only imagine how much more you could get out of it if you had a decent aftermarket CPU cooler!

__________________

twisted_58

nice job. get a better cooler and you could prob get that thing to 4GHz!

__________________

We Are Gr33n Machine

4 GHz Overclock Club

highoc

Im not that rich i live in Sri Lanka here there is no aftermarket cooling stuff

and i dont have a credit card not even my dad.'

__________________
[/CENTER]

UkuleleGod

Nice work. I had to put 1.40 volts onto my chip to get it stable at 3.01ghz.

has it been linpack/OCCT/prime95 tested?

__________________

TwoCables

If it hasn't, then it's imperative to make the overclock stable enough to do this:

• Approximately 12 hours of the Small FFTs test in Orthos (Orthos is Prime95, but it has a better GUI).
• Approximately 6-12 hours of the Blend test in Orthos.
• Approximately 5-10 passes in Memtest (1 pass is when Memtest completes all 8 tests one time).
• Approximately 30-50 linpack runs using LinX. This will get the temps 10-15°C warmer than the Small FFTs test. But it's pretty safe since it's only for an extremely short period fo time (less than several hours, that is). Be sure to click the 'All' button next to "Memory to use" so that you get the most out of the linpack test.

If the overclock is stable enough to do all of this, then you can definitely rest assured that you will never run into problems due to instability.

Now, you may have noticed that I didn't mention OCCT. This is because I think it's pointless to use OCCT with the existence of Orthos and LinX. For one, both Orthos and LinX open with in a fraction of a fraction of a second. Yeah, I meant to repeat that. It's just that fast. But on my system, OCCT takes 12-15 seconds to open, and I think that's absolutely ridiculous. If I want software that takes 12-15 seconds to open, then I'll go build a legacy system with a 200 MHz Pentium, 32MB of memory, and onboard video as well as an 8.4 GB 5400 rpm IDE HDD.

I mean, the Small FFTs test is the very same test as the Small Data Set test in OCCT. The Orthos' Blend test is the very same test as the Medium Data Set in OCCT. The linpack test in LinX is the very same linpack test that OCCT has. But the thing I hate about OCCT is that I feel like I'm running blind when I run its linpack test. I don't get to see how many runs it has done, nor does it give me any indication as to whether or not it's stable. The only indication I get is whether or not the system locks up or blue screens. Not only that, but I don't have nearly as many options in OCCT as I do in LinX. So, since OCCT takes so ridiculously long to open, and since I have the same tests in other programs, I see absolutely no reason to use OCCT.

And regarding the Orthos/Prime95 thing: the creator of Orthos took Prime95 and gave it a new GUI, and then he also stripped it of its useless features (that is, they're useless to overclockers). After that, he called it Orthos and released it. Otherwise, it's no different than Prime95. Actually, there is one difference: if you want to stress all 4 cores of a quad with Orthos, then you have to run two simultaneous instances, and then right-click on either instance and set the Affinity so that one instance is stressing like cores 0 and 1, and the other is stressing cores 2 and 3. Orthos does not natively stress a quad like Prime95. But that might be because Orthos was created before quads came out.

One last thing: if a 12-hour stress test seems like an impossible amount of time, then run the test while you're sleeping. Start the test on a night when you have to wake up for school, work or something similar that ensures that you will be away from the computer for at least 12 hours from the start of the test. However, before going to bed, be sure to monitor the temperatures for the first 5-10 minutes so that you know what the maximum temperature is.

__________________

highoc

i ran Orthos abt 1min

im scared to goo too far because my mobo dont have warranty

__________________
[/CENTER]

TwoCables

But like I said, if you don't ensure that the overclock is stable, then you could be frustrated by problems due to instability. This is because every overclock is unstable by nature. That's why servers are never overclocked (Overclock.net's servers are not overclocked). I mean, a server needs to be technically 100% stable 24/7. But for the way we all use our systems, we can make our overclocks practically 100% stable, which is good enough for daily use so that we never run into problems due to instability.

Besides, as long as your full-load temps with the Small FFTs test aren't exceeding like 80°C or something, then you have absolutely nothing to worry about. Besides, you probably already voided the warranty by overclocking.

__________________