PIII 700/100 @ 133, success , but unstable . SAD. Please I still need help.

[zulu]

Well, it seems that thinks are a little bit more complicated than I thought...
1. When I turn on the comp. it tooks him time to POST, or it don't POST at all untill I press RESET button. Then it loads Windows e.c.t.

2. I run 3dMark 2000 benchmark and demo without lockups, and the temp :
idle-27, load-32-34.
But 3dMark 2001 feezes after 3 min either in demo or benchmark (it isn't comprtibility problem , I run it before @ 112MHZ FSB.), and the temp. is still 35 degrees.
Sounds like what to you?
Voltage?
Overheat?
Memory?
Should I do Burn in?
What alternative to Seti and Prime95 do I have? Is running 3dMark 2001 loop demo enough?
What is Q3A command for loopdemo....
Thanks in advance.

[Phil]

what voltage is that at and what cooling? the boot up problem sounds like you have set a higher voltage, but with some boards the voltage isn't set until during the post, in which case if you need a certain voltage to boot it won't even post, but the resetting seems to restart it with the correct voltage, you may need up to 1.9v, but I don't recomend going over and I will say you shouldn't go over 2v, is it a cCO or cBO chip, it does sound like a lack of voltage though.
If you want to burn in I recomend setting it to 700mz with 2v and running prime95 torture test and sandra burn-in wizard, if you are using standard intel cooling though this may be a problem, I recomend the FOP 32-1, this is a good heatsink and the fan isn't too loud, if you end up needing more cooling but don't want the noise of the FOP 38 I recomend buying a 47cfm ys-tech 80mm fan and putting this on, this setup cools my cCO celeron 600@1.01ghz really well

[zulu]

Sorry,
Voltage bump: from 1.70 to 1.80.
Idle temp:27, after runing 3dMark2000 for 30 min - 34.
PC 100 128mb Micron memory.

[batboy]

Zulu, you never did mention what sort of cooling you're using. Sounds like that could be part of your problem. Once you do some cooling mods, then install an excellent CPU heatsink and fan using good quality thermal paste, try again. If it's still unstable, try increasing the voltage another notch. The other thing that could be holding you back if your motherboard doesn't have a memory bus divider is your PC100 RAM. If you can borrow some PC133 RAM, you might try that. Good luck.

[zulu]

Well I went back to 112 FSB (it was stable as rock before, really) and run again 3dMark 2001, and it locked. Still 3dMark 2000 runs for 9hours and the temp was 34 degrees! So WTF? Could this little voltage bump to 1.8 with normal temp,still derange smth. in CPU?

[zulu]

Well I went back to 112 FSB (it was stable as rock before, really) and run again 3dMark 2001, and it locked. Still 3dMark 2000 runs for 9hours and the temp was 34 degrees! So WTF? Could this little voltage bump to 1.8 with normal temp,still derange smth. in CPU?

[Lancelot]

I noticed you have PC100 memory, make sure it is set to cas 3-3-3 in your BIOS setup, cuz even if it is PC100 cas2-2-2 stick, it prolly won't run cas 2 at 133Mhz...

[logos]

have you burned it in? seems like your temps are fine. like phil suggested, knock it back down to 100fsb, up the voltage and BURN IT IN.

also, as your temps are good you might think about scoring pc133 before upgrading hsf

[zulu]

I'm burning it @112 FSB, and 1.8 Vcore. Is it enough?

[logos]

you can probably go higher- just be watchful of temps. with this chip i burned in at 1.8 for while, then 1.85, then 1.9 because it ran too hot. finally it ran cool even at 1.9 (max w/ this board) right now I loaded a program to run the voltage at 2.0 for more burn in...

[zulu]

I dont want to go higher, I want it to be stable @ 133 with 1.8. So I should burn it at higher voltage than I'm going to?

[logos]

yes, generally you burn it in higher than it will eventually run, but if don't do anything you don't want to do. i.e. the chip in this box wouldn't really go above about 90fsb until I ran it at 66fsb/1.8v and then 1.9 v, but then after burn in it ran over 100fsb fine back down at 1.8v (default 1.65)

every chip is different, I have built machines where a 600 or 633 cely will go right up to 100fsb without a v change or burn in. it just happens that the one I have took a LOT of work. maybe yours will too....

[Phil]

zulu (Apr 20, 2001 01:32 p.m.):
I dont want to go higher, I want it to be stable @ 133 with 1.8. So I should burn it at higher voltage than I'm going to?

I recomend setting it to 700mhz(7x100) and 2v for burning in, but in my experience burning in only allows you to run at a lower voltage, for example if you needed 2v to get a certain speed stable burning in may allow you to get that down to 1.7-1.8v or so.