my p4 ocing results...some good some bad

[Duvie]

First,

I had formatted NTFS, and I couldn't read my files on my other harddive which is in FAT32...Anybody know a way around this without having a dual boot system???? I couldn't get to all my good utilities to try like 3dmark2k1....

Secondly,

I had some old revisions of programs on backup disk and I had some compatability issues...

1)wcpuid version 3.0...doesn't read any numbers correctly or the ram speed at all

2)sandra 2001 read some things but wouldn't list memory speed or pci speed past like 110fsb...

3)Msi's fuzzy logic doesn't read anything correct either....Msi's pc alert reads but temperatures vary 4-6c depending on program...


So here we go....


Got my 1.8a to post at 2.4ghz...ran a bunch of the sandra test and some winxp programs...no freezes...played w/ memory timings and settings but got scores worse then my best which was at 122fsb...Is this a sign the ram can't take the speed....
I also when restarting from windows only would get the monitor to turn off and I would have to hit reset switch...


here are speeds and vcore voltage I used...


100fsb for 400fsb w/ p4 I got 968/1010 with sandra 2001....vcore set to 1.5v but actual was stated by several progs as 1.44v

110fsb for 440fsb w/ p4 I got 1048/1099....vcore set to default 1.5v...same actual reading

both above I saw temps from 41c to 45c

120fsb for 480fsb w/ p4 I got 1150/1199....I bumped vcore to 1.55v for actual 1.52v and it seemed to take care of restart issue...

122fsb had some restart issues, but played with ddr voltage and ap voltage and think I took care of it...
I got best memory scores here with 1169/1219...I tried vcore 1.575 for actual 1.54v

125fsb I had no issues at 1.55v which makes the above seem weird..I restarted several times tweaking memory and no issues...
My memory scores got worse though 994/1033...ddr voltage set to 2.6 and agp to 1.6v

temps started reporting anywhere from 46-49c with no real load....concerned...but I will try MBM when I get the other harddrive files to recognize.


GOOD NEWS???

It posted and started into windows out of bios at 133fsb for 533fsb and 2.4ghz...I ws able to run all the same test, only issue was the restart issue again....I boosted it up to 1.6v for an actual 1.54v to do this and even tried 1.625v for an actual of 1.57v but it didn't solve the reboot issue. It would restart but hang at MSI post screen which makes me think the ram may be the issue here. However since no prog reads correctly I couldn't see if at 133fsb the pci divider was set to 1/4...ram score went back to 1061/1101 which were better then 125fsb...so maybe....

Temps I feel were too high even if being reported differently...I did not want to boost the voltage any higher until I try AS3 instead of the thermal tape on theretail heatsink...



Thoughts:

1) Why does MSI report sctual readings lower then set vcore...My sparkle 350watt PS always seemed to add a few to athlon vcore?

2) Diminishing returns after 122fsb which could have bee equal to 155mhz or 310mhz for the ddr-ram and maybe my crucial pc2100 can't do it...Maybe some pc2400 is in the works???

3) reboot issues may need more voltage but in some reports I have heard on other boards all the voltage increase in the world wont help as it is the board that is not applying the right voltage at bios boot-up. Does this sound right?

4) I would like to apply AS3 to p4 and chipset hsf and see if I can get the temps down 3-4c...

5) May reinstall winxp at fat32 so I can access my other drive and get the real hardware testing programs....


Conclusion:

I am happy if I can get it to stay stable and better temps at 122fsb for the better mem scores and 2.2ghz with a good safe vcore... If you think that I can hold the 2.4ghz please advise me on what you would do to remedy heat, power, and restarting issues....

 

[KenAF]

duvie,

<< It posted and started into windows out of bios at 133fsb for 533fsb and 2.4ghz...I ws able to run all the same test, only issue was the restart issue again....I boosted it up to 1.6v for an actual 1.54v to do this and even tried 1.625v for an actual of 1.57v but it didn't solve the reboot issue. It would restart but hang at MSI post screen which makes me think the ram may be the issue here. However since no prog reads correctly I couldn't see if at 133fsb the pci divider was set to 1/4...ram score went back to 1061/1101 which were better then 125fsb...so maybe.... >>

I don't know the cause of this reboot issue, but I know it is very common on several boards (MSI, Abit) using the the SIS-645 chipset. Those on other forums think that it will be solved with a bios update. It has nothing to do with your processor's ability to run stable at that speed, nor does it have anything to do with your memory, AFAIK; it's an issue with the motherboard. The only SIS board that I haven't seen the rebooting problems mentioned for....is the Asus P4S333, but then, I have not looked carefully. I do know that the Intel-based boards from Asus and several others do not have this problem.

If you can't find a workaround for this motherboard bug, then you may have to stick to <2.4GHz until MSI releases a bios update to fix the problem (I'm assuming that they will, since so many people report this problem on the MSI and Abit SIS boards).

As for your Sanda and CPUID programs...it sounds like you are using very old versions, so its hard to compare.

 

[THUGSROOK]

what ratio are you running your memory at? 1:1, 4:5, 3:4?

run your memory at 1:1 to remove it as a limiting factor while overclocking.

(thats why i run all those weird tests - to single out individual overclocking limitations)

 

[Duvie]

There is no dividers I have been able to find for the memory...

There is 3 options in memory....200mhz which could be distinguished as 1:1...266mhz which could be 4:3...and auto which I don't have a frekkin clue...

Also there is spd, cas 2 and cas 2.5 for ram timings....I tried most all of them....


One question how does anandtech say they get 140 stable??? Does he think not be able to restart the computer is stable?? That is frekkin lame!!!!

 

[KenAF]

Duvie,

Never, ever take the word of one review, or even several user reports...

In the future, search the major forums like Arstechnica.com, Hardforum.com, here, and manufacturer-specific forums to see what real world users (people that are runnin systems for more than a few hours) are able to get.... The rebooting problem with the MSI has been mentioned on this forum many times before.

 

[Duvie]

I have actually researched it will and read the review at www.ocworkbench.com when the said they ran it up to 176fsb stable with an unlocked cpu...I didn't just read one...

But f^ck if these idiot sites think that restarting issues are stable... These sites should have faced this issue.......


One thing I don't seem to have issue at 125fsb...I restarted about half a dozen times....I have had it at 122 but tinkered a nit and made it stop....I have only tried 133 above that but felt it was fruitless as the memory was vastly diminshing...Is this potentially errors in sandra or were these real?

 

[KenAF]

Duvie,

I presume you've compared memory tests with the 200MHz (probably 1:1, as you suggest) and 266MHz (probably 3:4, as you suggest) settings? At 3:4, 125FSB would put the memory at 166MHz...can your memory do that much?

 

[Duvie]

yes I did...I did notice at 133fsb the memory scores looked a lot like my 266mhz memory ones at 100fsb default...I think the pci divider of 1/4 kicks in there but I can't tell until I get to some of my utility programs....



What do you think about the heat and or the reported temperature differences by all the softwares??? I have MBM 5.0 I can install but I wonder if I need a new downloded version since the northwoods are so new....

 

[THUGSROOK]

hiya duvie,
1/4 pci should kick in at 121fsb.
set your memory cas to auto.
set your memory ratio to 1:1. (your memory flaking out isnt gonna help us at all)

ignore software programs completely. the most accuate reading will come from the bios, and even that may not be right! only trust software if it agrees with what your bios says. the msi software will prolly work the best.

is your USB enabled? try disabling it to see if that helps. disable anything in bios you arent using too. (serial1, serial2, printer port, etc)

 

[Duvie]

I already disabled eveything onboard except for the usb stuff I am tring to use the usb 2.0 stuff...I will try disabling that stuff...I haven't plugged any usb devices in though...


I think the software needs an updating considering there own software doesnt f^cking work...It may be a northwood issue which may lead to the assumption it needs a bios fix and maybe the reason why anandtech and ocworkbench didn't mention the reboot issue eventhough they clocked the fsb higher then I have. They also likely oc'd the fsb on a special unlocked p4 thus needing no voltage modifiction of the vcore...I don't know speculation at this point...