GA-EP45-UD3P - Won't boot with 450FSB

[D-an-W]

Re: GA-EP45-UD3P - Won't boot with 450FSB

I've had this issue too with OCCT on Win 7. I also assumed that the graphing module/section of the program is the part having trouble. A 24 hour Prime of the same sized data set revealed no errors, which supports this assumption.

Since then I have always run OCCT in Vista SP1 compatibility mode, and have never experienced the error since. Compatibility mode, as far as I can tell, doesn't impact on the stability testing aspect of OCCT and it carries on doing its thing to full effect.

Just done ANOTHER 24hrs of OCCT Medium with it set to Vista SP1 compatibility mode and as before it stated it completed without error but Windows reported a problem at the point where the images etc are made.

I ran a 9.5hr test previously with the same settings and it worked ok, I fancied OCCT as you can tell it to stop at 24hrs but it's back to Prime I guess...

[Octopuss]

Re: GA-EP45-UD3P - Won't boot with 450FSB

I am 14 hours into Small FFTs with 1.33750 VCORE (540x7.5).
Damn I can see my ultimate achievement: 48 hours of my custom blend test. See you in a year :D

[profJim]

Re: GA-EP45-UD3P - Won't boot with 450FSB

I am 14 hours into Small FFTs with 1.33750 VCORE (540x7.5).
Damn I can see my ultimate achievement: 48 hours of my custom blend test. See you in a year :D

In your dreams

[Octopuss]

Re: GA-EP45-UD3P - Won't boot with 450FSB

I hate you!!
Really huge orange-blue toad will come lick your deepest during the upcoming nights!

[Octopuss]

Re: GA-EP45-UD3P - Won't boot with 450FSB

Damn thing failed after 18 hours!! Aaargh!
But at least it looks like CPU stability is not dependant on FSB frequency, because I had to use 1.34750 VCORE with the 450 FSB template as well!
Question: How higher than normal would you go with CPU ref (in case I wanted to try some desperate tweaking to get lower VCORE stable after all).

The RAM gains are purely from your FSB over clock. Notice how write operations gain a huge amount and read/copy only gain around 500MB-1000MB.

Just spotted this. How do you mean, just from FSB overclock, when the resulting RAM frequency is identical?

[Psycho101]

Re: GA-EP45-UD3P - Won't boot with 450FSB

Damn thing failed after 18 hours!! Aaargh!
But at least it looks like CPU stability is not dependant on FSB frequency, because I had to use 1.34750 VCORE with the 450 FSB template as well!
Question: How higher than normal would you go with CPU ref (in case I wanted to try some desperate tweaking to get lower VCORE stable after all).


Just spotted this. How do you mean, just from FSB overclock, when the resulting RAM frequency is identical?

For CPU ref, you'll be unable to boot long before you set a high enough voltage to damage anything, that's assuming that |ref volts can damage things.

As yo increase FSB, synthetic RAM benchmarks increase as there's more physical bandwidth between CPU, MCH and RAM in both that direction and the oposite RAM, MCH, CPU direction, leading to better scores. Because of this it's easy to conclude that 333*3.2 is lagging behind 533*2 because it isn't synchronous. It's not synchronysity that's doing it, just a faster FSB.

It's also very difficult to compare synchronous and asynchronous as obviously one can't set the same CPU frequency, same FSB and the same RAM speed and have both a synchronous and an asynchronous config to test, it's impossible. Hope that's clear on what I was trying to say. I think I've confused myself looking for the best way to explain.

[Octopuss]

Re: GA-EP45-UD3P - Won't boot with 450FSB

Luckily 450x9 is same as 540x7.5 in both ways :D
I agree on the arguable part of benchmarks. It still gives you a general idea though...

edit: higher CPU ref might help me get one notch lower VCORE stable or something like that, right? But probably not more. But how much higher than normal still serve any purpose? I usually try to lower it once, if still unstable twice, but not more.

[Octopuss]

Re: GA-EP45-UD3P - Won't boot with 450FSB

I believe in toads now. Looks like I gotta raise MCH core AGAIN, this time to 1.32, else no stability for me. Hate it!!

[Lsdmeasap]

Re: GA-EP45-UD3P - Won't boot with 450FSB

Pretty soon it will need 1.4 if you keep stressing your system 24/7

[Octopuss]

Re: GA-EP45-UD3P - Won't boot with 450FSB

I don't get it really. How is it possible the MCH voltage requirement obviously goes up? I know, with a fan on the new NB heatsink it's not too much of a problem, but it pisses me off from general principle!
What's worst - I raised MCH core to 1.32, BUT thought 540 FSB is too low, so I also went up to 550 :D Gonna give it a shot. lol. It's neverending.

[Octopuss]

Re: GA-EP45-UD3P - Won't boot with 450FSB

Bleh. Last spam for a few days.
550 doesn't work :D
Let's try 1.280 MCH core and 534x7.5! It's about the VCORE again. 1.32 (maybe lower!) vs 1.36 for 50MHz gain? Doesn't work for me.

Wish me luck your warriors.

[D-an-W]

Re: GA-EP45-UD3P - Won't boot with 450FSB

Finally nailed 510x8.5

24hrs each of Blend, Large and Small (Did some heavy work during each test as well). Will post the final template tonight, gotta work now...

[profJim]

Re: GA-EP45-UD3P - Won't boot with 450FSB

Nice!!!!!

So how soon will you start testing with 520 x 8.5?

Has your kitchen and bathroom overclocks passed 24 hour stress testing?

[D-an-W]

Re: GA-EP45-UD3P - Won't boot with 450FSB

Morning PJ!

The Induction Hob needs reseating before we can use it!!!

Might have a quick look at 520 tonight but I think temps will be getting above the comfort zone if much more vcore is required.

[profJim]

Re: GA-EP45-UD3P - Won't boot with 450FSB

Just make sure that you use a high quality TIM for the Induction Hob (sic).

I recommend that you rest on your laurels for at least 1 day before you tackle the 520MHz tweaks.