Gigabyte GA-P35-DS3L Thread

[Coelocanth]

Originally Posted by Javy

My rig is:

CPU: E6300 (OCed @3.10GHz)
RAM: DDR2-667 G.skill 1GBnj in Dual Channel mode (SPD 2.5= 5-9-9-25)
MOBO: Given right

Vmods:
Core-1.375
DIMM-.03
FSB-.01
MCH-.01
PCIX-.00

When I go to increase my Bus Speed to lets say 480 (3.36Ghz) there is a huge vmod jump ~1.48. My processor gets kinda hot ~60C at load. After seeing SGTSpikes OC (3.3 @1.26V) I am wondering if I am doing something wrong.

I have tried just one stick of RAM with an SPD of 2 and my PCIX is set to 100. Is it my memory DDR2-667 or something else like powers supply?

Thanks for the help in advance.
My 3dMarkScore: 12603 (link)

Every chip is different, even from the same series/manufacturing run, so ther's no guarantee that you'll be able to get the same FSB and vcore values as someone else. As for the vcore jump when you raise the FSB, that's to be expected. As you go for a higher and higher bus, your chip will need more and more juice for each successive jump in bus speed. Your temps also rise exponentially as vcore rises, so that's why you're pumping more heat.

But it seems from the way you worded this that you've got your vcore set on auto. Forgive me if I'm wrong, but is this the case? If so, go in there and set it manually. The board will feed your chip far more juice than it needs to stay stable, so you may find you can drop the vcore and still get a stable FSB of the same value.

[bioniccrackmonk]

Temperature questions for OC'ing a computer;

What is a good up to date program to use for monitoring temps? - the Easytune only tells me the CPU and Case Ambient temps, is there other items to watch?
What temp would be considered high for an E8400?
What temp should everything stay under during full load to be considered safe?

I have already started to OC my system, very slowly but surely, and I already see an increase in temps but I don't know what would be high for them. thanks for any help.

[Javy]

Originally Posted by Coelocanth

Every chip is different, even from the same series/manufacturing run, so ther's no guarantee that you'll be able to get the same FSB and vcore values as someone else. As for the vcore jump when you raise the FSB, that's to be expected. As you go for a higher and higher bus, your chip will need more and more juice for each successive jump in bus speed. Your temps also rise exponentially as vcore rises, so that's why you're pumping more heat.

But it seems from the way you worded this that you've got your vcore set on auto. Forgive me if I'm wrong, but is this the case? If so, go in there and set it manually. The board will feed your chip far more juice than it needs to stay stable, so you may find you can drop the vcore and still get a stable FSB of the same value.

Thanks for your reply:

I have manually incremented the FSB and Vcore accordingly. While increasing these quantities I found them to be somewhat linear and I am able to follow. But I seem to hit a wall for a Bus Speed > 459. My findings were as follows (20Hz increments I know that is kind big but bear with me)

Stock:
Bus speed= 266 Vcore=1.35000

OCed:
Bus speed <= 459 Vcore=1.35625 (not bad for OCing 200 from stock)
Bus speed = 460 Vcore=1.41875 (notice the significant jump)
Bus speed = 480 Vcore=1.48760 (another jump)

So I started to think what could cause this. And wanted to get some feedback. Maybe unstable PSU or bad RAM?


Below is a thought experiment, followed by two actual experiments. I invite everyone to constructively criticizes and help develop:

Thought Experiment:
If I were to make a conjecture, it would be that there is a far greater variation between different manufacturers of power supplies than the quality control yield of CPU and MOBO from the same manufacturer. I could see a power supply reaching some sort of threshold where the voltage begins to considerably unstable relative to the sensitivity of the CMOS transistors on/off threshold values inside of the processor. This voltage would again stabilize at a certain increase, hence a jump in the Vcore would make the system stable again. As transistors reduce they require less energy, (ie potential differences and current) to switch, so PSUs stability, not the actual rated power (ie 630W), becomes critical. The stability would come into play via the building process, quality of materials, engineering, and quality control of the manufacturer. So even though you get a cheap PSU you could have an oddity that is extremely stable at the voltages necessary for a processor.
Same could be said for the PSU relationship with the RAM instead of the processor.

Experiment 1:
So I have another PSU but it is the older ATX spec, 20pin, P4 cable, etc. But the system will still boot with this PSU and I have all the necessary cables to supply power to all the peripherals. Maybe I should plug this bad boy in and see if the Core voltages change or become more stable etc.

Experiment 2:
I also have another stick of ddr2 ram from a Wintec but it is DDR2-533 512MB. I could plug this in and see if there are variations in the core voltages with just different ram.

What do you guys think? Should I carry out these experiments or has someone here already done this and save me the time.


Then you have the timing issue of the RAM and CPU this I am much less familiar with and probably need some help on but one thing at a time.

Thanks,

Javy

[dleccord]

Why isn't this thread stickied yet?

[c230k]

Originally Posted by bioniccrackmonk

Temperature questions for OC'ing a computer;

What is a good up to date program to use for monitoring temps? - the Easytune only tells me the CPU and Case Ambient temps, is there other items to watch?
What temp would be considered high for an E8400?
What temp should everything stay under during full load to be considered safe?

I have already started to OC my system, very slowly but surely, and I already see an increase in temps but I don't know what would be high for them. thanks for any help.

Well it seems many people have an idle temperature of around 41 regardless of what cooling they got. I am not sure if it's a bug or what coz even mine is like this. For under load temperature, Intel said 72C is the max u would want to go, but I think it should be best to stay under 60C. Mine is under 50C under 100% load. Hope it helps.

[bioniccrackmonk]

Originally Posted by c230k

Well it seems many people have an idle temperature of around 41 regardless of what cooling they got. I am not sure if it's a bug or what coz even mine is like this. For under load temperature, Intel said 72C is the max u would want to go, but I think it should be best to stay under 60C. Mine is under 50C under 100% load. Hope it helps.

Thanks for the reply. I am not sure what "bug" everyone is referring to, but I am checking my temps with the EasyTune Pro 5 and it says my CPU idle is 17/18C and I really haven't checked it under load yet. Is anyone using different software for monitoring temps? If so, any recommendations are greatly appreciated.

[dleccord]

i use core temp .96 and i idle around 20-22C

[Coelocanth]

I use CoreTemp as well and my idle temps are 26C, with load temps at 57C.

[skim21]

Originally Posted by Splinky

hey c230k,

do you use stock NB cooling? how are your voltages to reach 500 FSB?

I heard the NB gets mad hot.

Thanks

Hey Splinky,
how did you overclock your e6550 to 3.0? I can only get mine to 2.8.

[skim21]

i got a E6550 with this board and Crucial Ballistix 2gb ram.

this is my first time overclocking and i've gotten it up to 2.8.

any way i can get that up to 3.0 or higher?

also im noticing that my speeds go down to 2.4 when idle and 2.8 when in load. whats up with that?

im getting these readings with CPUz.