AthlonXP: Overclocking via FSB

[clicknext]

I want to overclock my AthlonXP 2000, but don't want to go through the tedius and possibly dangerous proccedure of unlocking it. So, I want to see if I can overclock it very far by FSB. My goal is to be able to increase it to 166 FSB so that the 1/5 divider takes effect. Right now, with only retail stock cooling, it can hit 2200 speeds and go through an hour of stress test stably under 60 degrees. I realize that's pretty high, but I'm going to to be getting a Volcano7 (and bending the thermal diode so that it touches the heatsink so that it has a temperature closer to actual) and increasing my voltage from 1.75v to 1.9v. The end result would be 2075mhz, from the original 1666mhz. Could it be done? What things might limit me? I can underclock the RAM seperately so I don't think that should be an issue, and the 1/5 divider would make the PCI/AGP a non-issue, right?

 

[EKAtBzboyz]

well i was looking at your computer specs for sky, the system in question, and i noticed a typo. its supposed to be pc2700 ram, just an fyi

other than that its only heat and the cpu itself that would be holding you back i would say.
if you wish,and the overclock doesnt seem to be getting to 166mhz, try pulling the gtxp, and then see if it could be that

also harddrives have been known to get flakey at high frequencies, i dont quite know how yours will hold up...if it gives a windows error before boot, thats usually the case

 

[chizow]

33mhz increase to the FSB is a stretch with such a high multiplier. Even if everything else checked out as you described, your cpu would have to handle the new clock speed. That much of jump in clock speed and voltage is gonna generate some ungodly heat on a Palomino, so you better have some intense cooling. Unlocked XP's or locked XP's with lower multis might yield better results, simply b/c the multiplier you are working with makes each increase to your FSB more difficult to narrow down any problems. 2ghz might be accomplished with a T-bred A, or a late stepping Palomino with some serious cooling, but AMD went to T-bred for a reason: b/c they had hit their clock/scaling ceiling around 1.8 ghz with the Palomino core.

Chiz

 

[clicknext]

Hmm, now I'm debating with myself about whether an overclock is worth it. Maybe ten to twenty more FPS in games? I guess I'm pretty much doing this because I think I can, not because I need or want the possible benefits, as with most overclockers. Perhaps I should just hold back and settle for a smaller overclock, like to 1.8ghz.

 

[chizow]

Heh, well OCing isn't something to be taken lightly. Don't take what people write here as gospel and just crank your FSB to 166mhz w/out changing the multiplier. At those speeds you can very easily fry your CPU and cause permanent damage to your mobo, components, memory, and HDDs. OCing is trial and error, thats why every bit of flexibility helps. You should test to make sure it is somewhat stable at 1.8 before even thinking to increase it. I personally like to establish what my CPU can do from the get-go so I don't have to spend extended periods of time tweaking, but if you are happy with your system's performance, no need to push it just to push it.

Chiz

 

[clicknext]

I just read this thread about changing the multi using a pencil, and it seems like this is something I could do, or at least attempt. Would it be much less troublesome to just up the multiplier? Is the effect on the CPU of doing this any different than upping FSB? Changing the multiplier only affects the CPU clock and nothing else, right? No danger of damaging other things?