• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Athlon XP 1700+ underclocking results

Leo V

Diamond Member
For those interested in minimizing CPU heat and building a near-silent PC, like mine with just one Panaflo fan. Here are my brief experiences with AthlonXP 1700+ (1466mhz).

I'll mention first that unlocking the XP was a rather tedious experience, because the conductive pen recommended by Tomshardware (store: Mouser) is extremely low quality--too poor to be able to follow Tomzilla's own instructions. Conductive grease would have been much better. Regardless, I'm happy that AMD at least allows unlocking the multiplier. If I was using Intel processors to build a quiet system, I'd be forced to resort to sub-par FSB speeds to reduce the CPU clock and lower the wattage. (P4's normally use unacceptably high wattage just like Athlons, as far as quiet fan cooling is concerned.) On to my results:

My torture test involved running BurnK6.exe alongside regular Windows apps. BurnK6 activates many parts of AMD processors simultaneously, requiring a near-maximum power draw. Thus, running at the default 1466mhz (133x11.0), the processor required full default voltage to avoid errors! This means it was using 1.71vcore and drawing a maximum of 64.0 watts. On my rig, socket temperature was approaching a dangerous 80C in this torture test, with case cover open. Max operating CPU temp is 90C for Athlon XP.

However, lowering the speed to 1333mhz (133x10.0) made a dramatic improvement. I was able to boot at 1.36vcore, but Windows crashed after loading. However, at 1.41vcore everything ran fine, and the torture test encountered no problems having run nearly an hour! By my estimates, this voltage corresponds to about 39.6 watts.

Just to be safe, I bumped up the voltage one more setting to 1.46vcore corresponding to 42.4watts max power. With the case cover closed as normal, and the torture test running for an hour, the socket temperature is only 67C--well within CPU spec.

That's a very nice achievement--1.33ghz and only 42.4watts. And I imagine higher-end XP's would do even better. It shows that the Palomino/XP core really is an improvement over Tbird. Well done AMD.
Leo
 
WHAT a slient PC? I didn't hear what ya said?😉

Nice post and very imformative
Can you post your sys specs? O/C anything else? ram ,vidcard, etc?
 
I've just updated my System Specs page (link in my sig), just follow the link to the first PC.

Nothing is OC'ed, I actually picked the vidcard for low heat dissipation (4watts on MX). But a GF3 Ti200 would probably work too. I used to be an overclocker, but the hair-dryer noise of the Air Peltier got to me 😉
 
definately remember that socket-thermistor XP temps are even more inaccurate than socket-thermsitor T-bird temps.... even on a higher reading mb, you may be 10+C under actual "die" temp, so 80C may well be pushing safe limits of CPU temperature... 67C should generally be fine, though.



Mike
 
I believe it should be possible, at this point, to build a 100% fanless AthlonXP system. There are a couple of companies making fanless PSU's, tkpower.com and (more recently) RSG Electronics. The latter have a 130watt psu capable of producing 230watts for up to 5 minutes, with 85% average efficiency. They are reportedly working on a fanless 350watt psu (good luck).

If you take a AthlonXP 2000+ (1.66ghz) and underclock to maybe ~1.2ghz at around 1.0Vcore (not too far-fetched), it would consume under 20 watts and could be passively cooled. And the power consumption should be within the fanless 130watt PSU's ability.

I'm going to contact RSG out of curiosity, but they're in Germany so it's not cheap to obtain their supplies.
 
Back
Top