Overclocking Stats and settings

riptide

Whats the most anyone has overclocked a northwood 3.4? What settings were used vcore, vdimm, fsb dividers, ram type etc.

I'm having trouble getting past the 3.75 mark @ 1.625vcore. I don't really want to go too far past the 1.7vcore mark, unless people have done it safely before.

Some specs are
Abit IC-7Max3
2 x 512 DDR500 Crucial Ballistix
600W EZcool PSU
Cooling is water - Waterchill CPU cooler
Processor is at 1.625v, FSB is 220approx, ram divider is at 5:4 (although the ram is hardly the problem it does 2.2.2.5 at 200FSB and is rated to 250FSB @ 2.5CL)

My instinct is that the northwood is at its limit at 3.8 or thereabouts, but, I know some have broke the 4Ghz barrier and beyond.

Actually maybe this thread can be used as a sort of database for people to add their settings and overclock reached.
Remember, it would be good if results are honest. I find on other forums that some results are outrageous.

DemonOfTheFall

I wouldn't go above 1.7v with a northwood. Isn't that dangerously close to where they can just suddenly die ? (SNDS - sudden northwood death syndrome)

riptide

DemonOfTheFall wrote:

I wouldn't go above 1.7v with a northwood. Isn't that dangerously close to where they can just suddenly die ? (SNDS - sudden northwood death syndrome)

Good point. The electical limit I've seen somewhere of a northwood I believe is around the 1.75 mark. After that its the 'danger zone' big time The SNDS I think was common enough with early northwoods, and also with using the likes of vapochill coolers. But I could be corrected.

DemonOfTheFall

Have no idea mate, never owned a P4. Just passing on a warning I've heard before. I believe it was 1.8v that was killing them dead in minutes, but 1.7v is a little close to that for my liking

ZiabR

DemonOfTheFall wrote:

Have no idea mate, never owned a P4. Just passing on a warning I've heard before. I believe it was 1.8v that was killing them dead in minutes, but 1.7v is a little close to that for my liking

Agreed, Tbh i would not go any further than you already have mate. Your at 1.625 atm and have achieved a nice OC so really no point in risking the chip.

gline

i got a 2.8ghz northwood to 3.4ghz stable @ about 1.8vcore on watercooling
You should get to 4ghz no problem with it, 1.75-1.8vcore "should" be ok

riptide

gline wrote:

i got a 2.8ghz northwood to 3.4ghz stable @ about 1.8vcore on watercooling
You should get to 4ghz no problem with it, 1.75-1.8vcore "should" be ok

I know I remember one of your posts before about some SERIOUS overclock Crazy stuff:D
As a matter of interest how long did you have the 2.8 @ 3.4 for ?

gline

riptide wrote:

I know I remember one of your posts before about some SERIOUS overclock Crazy stuff:D
As a matter of interest how long did you have the 2.8 @ 3.4 for ?

it was like that for about 6-9 months then my mate upgraded to a 3.2ghz prescott (it was his machine i overclocked)

unkel

It's a while I was playing with the P4C

riptide wrote:

My instinct is that the northwood is at its limit at 3.8 or thereabouts

You might be right, but you'd be unlucky to have a chip that would only go that far. Most go further. The cpu is the only potential bottleneck in your system for going higher (although I wouldn't trust that PSU)

Easy to try how far you can go: up the vcore to 1.7V, get rid of your divider (not good on P4C - only use if you have to) and set the ram to 250FSB @ 2.5CL

Up the FSB in small steps and run pifast / superpi until you get instability. Take a few steps back and check stability with prime95

DemonOfTheFall wrote:

I wouldn't go above 1.7v with a northwood. Isn't that dangerously close to where they can just suddenly die ? (SNDS - sudden northwood death syndrome)

SNDS applies to P4B, not P4C. But you're right, 1.7v on air is generally accepted as safe for permanent running. Higher on H2O. Check the temps though!

riptide

You might be right, but you'd be unlucky to have a chip that would only go that far. Most go further. The cpu is the only potential bottleneck in your system for going higher (although I wouldn't trust that PSU)

Ya I might be unlucky. Well the rock stable max is 3711.99 / 218.35FSB. A shade short of 10%. Ramping vcore did no favours and on a 'rosary beads in hand's effort at 1.8vcore ONCE- didn't even post! 1.775 won't bring it above the result shown. At 3.8 she will boot but hangs dreadfully at the log on screen and after 30 seconds frozen it will continue to desktop. But i get the odd windows error message and zone alarm gets anal.
Therefore, since I need rock like stability, the tools are put away and 3711.99 is where it'll remain at 1.6 vcore.
I'll keep the eyes open for another northy and an ES.

Do keep posting up your own experiences and your own results. AMD results are also welcome.

unkel

So did you try running the ram at 1:1?

Also, if stability is a main concern, loose the PSU and replace it with a decent one. 600W means nothing in a low end PSU

riptide

So did you try running the ram at 1:1?

Also, if stability is a main concern, loose the PSU and replace it with a decent one. 600W means nothing in a low end PSU

I did indeed. And 5:4. And i had the AGP/PCI locked at 66/33.

I even got the Ram to work for a while at 2.2.2.5 @ 218MHZ & CPU @ the 3711.99 mark and started benching and corrupted the disk. Had to fix with windows restore/ 'fixboot' command .

unkel

riptide wrote:

I even got the Ram to work for a while at 2.2.2.5 @ 218MHZ

Forget about the divider mate! Post your score 1:1 and preferably with a half decent PSU

riptide

unkel wrote:

Forget about the divider mate! Post your score 1:1 and preferably with a half decent PSU

Yes mate at 1:1. Still won't beat the 218Mhz. I'll look into getting another PSU as a lend and see what happens.

unkel

riptide wrote:

Yes mate at 1:1. Still won't beat the 218Mhz

Your Crucial is CAS2.5 but are you sure you haven't it running as CAS2? Could be a coincidence, but the 218Mhz is very similar to the magical speed that TCCD memory doesn't seem want to go over @2-2-2-5

gline

running a divider isnt a big deal for a decent overclock

riptide

unkel wrote:

Your Crucial is CAS2.5 but are you sure you haven't it running as CAS2? Could be a coincidence, but the 218Mhz is very similar to the magical speed that TCCD memory doesn't seem want to go over @2-2-2-5

Unkle I had it on 2.5 most of the time. But i gave up on the CPU and started tightening the ram up. Got it to 2.2.2.5

Its rated at 2.5 up to 250FSB, and 2 at 200MHz. Apparrently they don't like 250+ much. Kinda does what it says on the tin and thats it. I thought it was good going for the sticks to get 2.2.2.5 at 218FSB all the same. Oh ya nearly forgot... they crapped out anyways at 218FSB at those latencies.....

SouperComputer

same core, but much lower clock.

I have a 1.8 northood @ 2.7 on air, 1.5V. Mobo has no provision for extra voltage so I dont know if it will go further. Im sure it will though.

Riptide, did you do your business with Davie_B yet?

unkel

riptide wrote:

Oh ya nearly forgot... they crapped out anyways at 218FSB at those latencies.....

So both memory @2-2-2-5 AND cpu top out at 218FSB?

Bit of a coincidence surely?

riptide

unkel wrote:

So both memory @2-2-2-5 AND cpu top out at 218FSB?

Bit of a coincidence surely?

I'm sure the memory tops out below this at 2-2-2-5 if I stressed it enough. I haven't checked how far I can bring the memory at these latencies. I can't get the memory above the FSB setting anyways. I can't get it to, say, 250Mhz and have CPU at 218Mhz

riptide

SouperComputer wrote:

same core, but much lower clock.

I have a 1.8 northood @ 2.7 on air, 1.5V. Mobo has no provision for extra voltage so I dont know if it will go further. Im sure it will though.

Riptide, did you do your business with Davie_B yet?

Souper

No reply to 2 PM's. Don't know what story is. I'm sure PM is working for me!

riptide

Well.. this thread is back from the dead. I did a ghetto style alteration to the water cooling. I partially froze 2 litres of water and but it in a container with some ice cubes... then took off the fan on the radiator and put the radiator in. Temps are hard to judge (Abit IC7-Max 3's are notorioius for having dodgy sensors) but I estimate temps were held at 25-30 at full load. I got her to max 4030 Ghz, at full load. No problems. After this, little windows erors started to appear... but hey at least I broke the 4Ghz mark, ghetto style! I'll investigate the idea of having this setup with the help of some refrigeration.. Anyone recommend something not as severe as phase, but a bit better than water...? or will it have to be DIY?

gline

riptide wrote:

Well.. this thread is back from the dead. I did a ghetto style alteration to the water cooling. I partially froze 2 litres of water and but it in a container with some ice cubes... then took off the fan on the radiator and put the radiator in. Temps are hard to judge (Abit IC7-Max 3's are notorioius for having dodgy sensors) but I estimate temps were held at 25-30 at full load. I got her to max 4030 Ghz, at full load. No problems. After this, little windows erors started to appear... but hey at least I broke the 4Ghz mark, ghetto style! I'll investigate the idea of having this setup with the help of some refrigeration.. Anyone recommend something not as severe as phase, but a bit better than water...? or will it have to be DIY?

waterchiller system, u need to insulate the cpu then anyway and antifeeze in the water. Phase is easier and better temps

riptide

gline wrote:

waterchiller system, u need to insulate the cpu then anyway and antifeeze in the water. Phase is easier and better temps

Yep. Phase may be easier, but costs a few bucks. And ya kinda figured that condensation would become an issue well below room temps. Antifreeze would have to be closed system. Smells bad does it not!

gline

riptide wrote:

Yep. Phase may be easier, but costs a few bucks. And ya kinda figured that condensation would become an issue well below room temps. Antifreeze would have to be closed system. Smells bad does it not!

well yeh it would have to be a closed loop, i thought thats what youd be doing?

riptide

gline wrote:

well yeh it would have to be a closed loop, i thought thats what youd be doing?

Well think about it. You could also but my example to extreme and run water / antifreeze through a radiator of sorts that was submerged in an antifreeze mixture that was sub zero and which was inturn cooled by refrigeration. The water in the loop is obviously closed, but the water surrounding the radiator would be... well in a freezer unit of sorts and well be more open... does anti freeze smell bad.? Somebody said it would!

gline

riptide wrote:

Well think about it. You could also but my example to extreme and run water / antifreeze through a radiator of sorts that was submerged in an antifreeze mixture that was sub zero and which was inturn cooled by refrigeration. The water in the loop is obviously closed, but the water surrounding the radiator would be... well in a freezer unit of sorts and well be more open... does anti freeze smell bad.? Somebody said it would!

most people just use a chiller for the water, rather then a submerged rad(dont think that would be as efficient)

riptide

gline wrote:

most people just use a chiller for the water, rather then a submerged rad(dont think that would be as efficient)

What type of chiller? A commercialy available one for the purpose? You'd be surprised about the rad! Instantanoues change in temps when I tried it!