LanParty + 2500+

[psykoikonov]

I just bought the above combo nForce II Ultra version A with a Barton 2500+ week 34 AXDEA. I'm still using my old RAM as budget does not permit new RAM till new year . I'm using 2 different sticks of DDR333 2.5,3,3,7 (Samsung and a no name with identical Samsung chips on it) in single channel for now. I started at default (166x11) and have been slowing working it up, stressing system hard as I go (back to back loops of PCMark2k2, 3DMark2k1se, 3DMark2k3build340, SETI@home. I got to 180x13 and crashed (using Vantec Areoflow). Had to reset cmos, 180x12.5 won't work now . So now running at 174x13. Default voltage when read from bios or WHM (shipped software) reads 1.61-1.63v, it ran at 2.19GHz on that. Couldn't get to 2.3GHz without 1.85v. Temps are waaaaaaaaaaay off, @ 2.3GHz and toasted for an hour it reads 35C in WHM and less in bios LOL.
1. Could the mobo sensor be too far away from CPU?? I didn't even look at it. I know that it was 30C in case (beside the north bridge) using digital thermometer, room temp 21C. Should I try MBM5?
2. If I adjust my ram timings from anything other than by spd (2.5,3,3,7) system will not detect ram, any ideas (I've tried 3,4,4,8 and 3,3,3,7)? Both modules have completed more than 20 passes of memtest at 180MHz by spd with no errors, they are ECC mdules but mobo doesn't have ECC.
3. If I raise the CPU to 1.85, and all others (Vdimm, agp, chipset) up by .1v is it going to hurt anything? I thought I read somewhere that +10%v won't do any harm, true?
4. How far can I take a 2500+ on air approximately? Also I read a review of this mobo and they couldn't get over 231MHz on fsb (using DDR500), anyone else find this? If this is the case then there's no point in me getting DDR500 is there? Perhaps they didn't increase chipset voltage?

Any comments apreciated.
Thanx
Psyko
<P ID="edit"><FONT SIZE=-1><EM>Edited by psykoikonov on 11/27/03 10:15 PM.</EM></FONT></P>

 

[pIII_Man]

1 try mbm5

2. some modules are just weird like that i have heard many people have this same problem with geil modules

3. As long as the cpu does not go over 55c full load you are fine...because you don't have any reliable means of checking temp yet i wouldnt go above 1.75

Ya most products are rated to run +/- 10% of their desired settings (normally can be pushed further) in short you should be fine.

4. 2.3ghz is the average...2.4-2.6 can be achieved.


If it isn't a P6 then it isn't a procesor
110% BX fanboy

 

[psykoikonov]

Thank you piii_Man.

After 10 minutes toast MBM5 reports:
CPU diode @ 50C
CPU socket @ 38C
Case Temp @ 30C (room temp 21C)

Idle MBM5 reports:
CPU diode @ 36C
CPU socket @ 31C
Case Temp @ 27C

I realize now that the hardware doctor's (software shipped) "CPU temp" is exactly the same as MBM5's CPU socket temp and hardware doctor's "Systemp" is actually MBM5's Case temp (northbridge temp right??), this explains the off temp of the other monitors and no real increase under stress (10C @ best).

BTW: this is @ 173.6x13=2257.2MHz 1.8V (CPUz), I'm certain I can run this at lower voltage, that just happens to be where I was at when I was running 177x13=2306MHz (reads 1.85v in bios adjustment but 1.80 in bios, mbm5, and hardware doctor). Voltages are very stable am using Enermax Whisper 430W.

Anyway is 50C getting to high? I took the core and applied thin layer of AS3 as per Artic Silver website instructions. Also it's winter here so room temp will be much higher in summer (30C+)...yikes.

Is it standard to need extra cooling/voltage for northbridge @ 225+fsb?

Psyko

<A HREF="http://forums.btvillarin.com/index.php?act=ST&f=41&t=421&s=205efc501496823a64ba7fbd37152ff2" target="_new">Psyko's Rig</A><P ID="edit"><FONT SIZE=-1><EM>Edited by psykoikonov on 11/28/03 00:04 AM.</EM></FONT></P>