Can't overclock as high in WinXP?

[Slickone]

I read someone saying they couldn't overclock their Athlon as high in Windows XP as they could in 98. Is this usually true?

 

[Spikesoldier]

Yes, to get matching stability from 98 to XP, you need to up the CPU voltage a tad.

 

[Wind]

<< Yes, to get matching stability from 98 to XP, you need to up the CPU voltage a tad. >>


U mean u need more juice on the CPU just to achieve the same O/Ced speed compare to Win98se ?

 

[nealh]

WinXP is more stable in some ways for me. My Athlon is running at 1380(10x138..1.85v same as with W98SE)...for instance if a program crashes I do not usually get a hard lock ..but explorer resets....the exception is MS Outlook 2002 occasional when it opens will cause a BSOD and the computer will reboot..the other "bad" program I have is Pagis Pro Millenium..it has caused some lockups..not sure whether this is a compatability problems or O/Cing problem

Yes he means WinXP needs more voltage to maintain stability for some cpus

 

[iamtrout]

Hi.
I'm the person who posted the XP and 98 thing. I finally got my athlon 1.333 stable at 1.518 by upping the voltage a whole .1V.
My thing's running at 1.89V and at 44C idle with all my fans on high. Jeez, this thing's hot, and noisy now because of the Delta. I think I'm gonna invest in a water cooling solution. I'm scared to think what my temps will be if I put the thing on full load.

 

[LanEvoVI]

Well 44 ain't that bad you know...full load should still be under or just at 50 deg. That is still in the "comfort" zone for most people.

 

[Wind]

<< Yes he means WinXP needs more voltage to maintain stability for some cpus >>


So, can I said tht the WinXP had the same capabilities in O/Cing as much as the Win98...just tht somehow or rather it just need more power to the CPU.

 

[Slickone]

But if XP needs more voltage than 98 to get the same speed, that isn't the same overclockability as 98. Since if you upped the voltage in 98 too, it might get even higher. ?

 

[Emo]

It's not that XP needs more voltage but it uses your system resourses better, therefore exposing any inherent instability of your overclocked system that might have gone unnoticed under Win 98.

 

[JesseKnows]

"Uses resources better" is an interesting way to put it.

My system (VP6, 2XPIII700@933) became less stable between Win2K and XP, had to increase Voltage from 1.75 to 1.8 (spec is 1.65).

My guess is that the system is truly idle less of the time, so the processor has less time off (in "halt" state) to conserve power, and ends up dissipating more power, hence the temp rises and stability diminishes. If that is true, then when fully loaded it should be the same - say with SETI or Prime95 running. Any direct experiments to prove or disprove this theory?

 

[CTho9305]

if either OS is unstable only when OC'ed, then your rig IS NOT STABLE! I've heard that NT kernels (nt, 2k, xp) do better idling than the 9x series, so they tend to be cooler, so I always stability tested in win98. Also... when using linux, make sure to boot with the NOHLT option when you stability test to disable software cpu cooling during idle. Basically.... if one OS is stable, but another isn't (yet is stable at other speeds) you haven't got a stable computer.

 

[Fox3613]

Indeed, some Operating Systems use resources in a more efficient way. I suggest installing FreeBSD for example; those overclocked system will crash even faster. Another thing: overclocking could even lead to data corruption on a long-term base, you just don?t notice it very fast with a Windows platform.