Which motherboard

i am buying a amd athlon xp 3200+ 2.2ghz, 400mhz..... what motherboard should i get to have the best performance possible?? thanx

The IT guy where I work just set me up with that same thing. He put it on the latest ASUS (is that spelling right?? ) mother board, and we had enormous problems with spontaneous reboots. Right out of the blue, the system would reboot. It turn out that the 400MHz front side was too fast for the board. When we clocked it down to 166 it worked (and still does). It won't run at 200Mhz, 166 max
kevin

What do you mean, 400 mhz was too fast for the board? I don't think it's the Mobo, it's the CPU or memory that can't handle the speed. Something must be defective or mismatched. I have a Barton 2500 clocked to 3200 speed. It's fine in my Biostar M7NCD. ASUS is a good brand. My Biostar was like half the price of the ASUS.

The MoBo is become a very weak link in a system because you have 64 signals that leave one place and need to arrive at the other place at the same time. To do this at 400 MHz the traces need to all have match impeadance, and if only one trace has more impedance because of crosstalk, grounding issues etc. then a problem will occure.
kevin

Hmmm ... anyone have the ohm ratings on these things? Guitar amps are 4 ohms, home stereos are 8 ohms ... motherboards are ?? Well, surely there is resistance and it must be a factor but I've never seen anybody use ohms to match computer components. Anyone want to enlighten me on this one?

I think you're just not using the right something. My (Asus MB) FSB is 800 and everything is working just fine.

Its not the real resistance measured in ohms the can kill a bus, it's typically the capacitance and to a lesser degree inductance (the imaginary parts of impedance) both have units of ohms when looked at in terms of AC signals. The impedances of caps and inductor is measured as follows

Zc = 1/jwC
Zl = jwL
where j = square root(-1)
w = the frequency of the signal

so are the switching frequcny goes up, so does the impedance of the bus.
As far as matching resistances, that is done to stop reflections of the line: if there is none of the signal reflected back, then 100% of the power gets absorbed by the load (speaker). It is easily kind of like seeing a reflection in a window. If the window reflects back all of the light on your side of it, you can't see out of it. When this is applied to busses, the longer it is, and the faster the switching speed on it, the more important it is to terminate it proberly so there are no reflections.
kevin

hope it helps

When I was a small child I discovered that if you plugged a microphone into a headphone jack it worked like a speaker.

As a late teen (18, 19 or so) I discovered that when I patched two electronic keyboards together in a radio shack Y cable and then into an input on my tape deck the sound from keyboard A did not, infact, merge with keyboard B and mix into the input but instead the sound of keyboard A ended up traversing keyboard B and somehow ended up coming out its speakers!

Then I took physics in college.

And that's why I'm so dumb today.

Actually that makes a lot of sense. But the only ASUS Motherboard I ever had a problem with was the one I fried, by accident, but it was definitely my fault. I've gone through several.

But I still think that



is a case of not using the proper parts somewhere in the chain.

I like the nforce based motherboards very stable... in linux at least

Here is a list of MB makers http://nforcershq.com/

ps:
My own workstation is based on Soltek frn2-rl and barton 2500+ ; Nforce2 Soltek is good at overclocking too

http://forums.amd.com/index.php?showtopic=692