The Official OC (OverClocking) Thread!

Hey, dont ask me. I dont think, I just do.

 

Lmao.
im: I'd also be curious to see exactly how the 6700 stacks up to the 9450 in a bench.

 

Only problem is that the pencil wears off after a while. So it's better to use conductive ink.

 

Then again, I've seen too many pencil mods (usually graphics cards) go up in flames, quite literally. Load line calibration like on the Rampage makes so much sense.

 

I'm not quite certain I understand your point. Do you mean that by adding a resistor the overall voltage is increased? That could be true if measuring total power draw resulting from using a resistor, but it couldn't it also mean a decrease in voltage exiting the resistor which could be desirable since the object is to protect ones processor rather than save power?

 

it's a nice feature but it can get abused too.

 

Pencil mods on GFX cards do different things. They increase the voltage so you can OC higher.

This makes sure that your voltage doesnt DROP.

 

My point is, why would you design a hundred-dollar plus motherboard that drops volts on the CPU under load, when it wouldn't if the value of the resistor was slightly different? I see why it works, but not why it was a problem that should ever have existed in the first place...

Mort: How so?

 

It's not their fault. I think it's something Intel implemented.

 

All designs are meant to protect the financial interests of the companies benefiting them and not the consumer. Restricting voltage in certain areas that might potentially sustain damage from increased voltage makes sense from an economical view, but they may be a little more protection than is needed. This is essentially what DFI Lan Party boards do with its overclock settings, because damage to their boards is probably factored in to operation costs.


I have got to start reading my sh*t before I hit save. LOL

 

Ahh I see, that makes sense now, thanks.

 

edited. Soph answered the question and better than me to boot.

 

A cinch huh?? I think that OC is pretty good buddy and besides the mIPS at almost 63K are a HIGH mark in this crowd. I see you got a full event with your torture tests. ;P

WTG buddy and GL hitting the next mark.....

.....gm

 

abuzar1,

If you say so, but vdroop is built in for a purpose on Intel motherboards to protect it and other components from line spikes and sudden voltage surges. All you are doing is cutting down or eliminating the amount of protection you get by having it. In effect you are eliminating protection that was purposely designed in. It can wind up being a very expensive mod!

sammorris,
You should set Load-line calibration from Auto to Enabled in the MIT section of your set-up! Here's the full story behind Vdroop that explains it all:
http://www.thetechrepository.com/showthread.php?t=126

Best Regards,
Russ

 

Lol so my board has it after all? I'm such a buffoon...

Thanks for the link Russ, but at 11pm it's all way over my head!

 

I've been doing some testing and comparing using RB/CCE SP and RB/HC as test plaforms. I used the exact same folder/dvd file for each test. all times include preparing, encoding, rebuilding and building an ISO image. I first thought HC encoder was faster than CCE but I forgot that I had CCE set for 3 passes. CCE with only 2 passes is faster than HC set at best quality. poor ole E3110 even at 4.0ghz was sucking hind teet. the results are as follows:

Q9450 @ 2.66ghz CCE SP 3 passes 51 minutes.

Q9450 @ 2.66ghz HC best quality 43 minutes.

Q9450 @ 3.4ghz CCE SP 3 passes 42 minutes.

Q9450 @ 3.4ghz HC best quality 37 minutes.

Q9450 @ 3.4ghz CCE SP 2 passes 33 miutes.

E3110 @ 4.0ghz CCE SP 3 passes 52 minutes.





I normally use CCE SP with 3 passes for all my encodes but since HC will utilize all four cores and is just a little faster (set at best quality) than CCE SP is using 3 passes, I might start using HC once in a while now.

 

Morty,
I don't know what size file you're using but knowing you it's about 7Gb or so.. Those times are impressive buddy... I'm very interested in getting my Q6600 back up and running. I've only got the psu to test next... it's NOT the RAM that was causing my issues... :-( *crosses fingers* in hopes it's NOT a psu problem.... I don't want to mess with an RMA on an Antec 1000 Quattro.... or the ASUS Striker for that matter...

.....gm

 

Mort81,

Correct me if I'm wrong, but I thought CCE was multi-Threaded. It uses both cores of my E6750 while encoding, or is that only for the Dual-Cores?

Best Regards,
Russ

 

it does use all the cores to some degree but nothing like HC does. Hc pegs out at 100%, all four cores.

 

Sorry for the delay in the response...

I will post the benchie a lil later.

But more importantly....I have a question:

My friend gave me an ASUS A8N-VM CSM motherboard socket 939.
Now, as most of you know, I don't have any AMD CPUs.
Would / Should the motherboard at least show me any kind of error screen or something if I were to power it up w/o a CPU in the socket?
Or do I have no way of checking that it works w/o a CPU?

(I wanna sell it on e-bay but I want to be certain that what I am selling actually works, if I can't test it....I won't sell it)

***I didn't feel it necessary to make a whole new thread for this but if you (Moderator or Master of AD) think I should move this post, I will comply.***

thanks,
-im1992