Extremely undervoltable Socket A board?

0Byte · Post by **0Byte** » Fri Aug 13, 2004 6:43 am

Hello Board,

does anybody of you happen to know an ATX Socket A board that is undervoltable to at least 1.00 V or even less? I have alread looked at the site about undervoltable mainboards in the recommended section. The only boards mentioned there that fit my requirement are supposed to be produced by Soltek. Having looked through all the BIOS Manuals at Soltek's website, I have, however, found out that all recent Soltek ATX mainboards for Socket A can only set the Vcore to a mere 1.10 V. Maybe this should be updated in the recommended page.
So again, do you know an ATX Socket A Board that is undervoltable to at least 1.00 V or is there at least an easy mod to archive that goal?

Regards,

0Byte

efcoins · Post by **efcoins** » Fri Aug 13, 2004 7:41 am

Which processor are you going to use at 1V ?

My AthlonXP is stable at 1.35V, with a reduced clock speed, but 1V, or 1.1V is useless.

sthayashi · Post by **sthayashi** » Fri Aug 13, 2004 7:49 am

efcoins wrote:Which processor are you going to use at 1V ?

My AthlonXP is stable at 1.35V, with a reduced clock speed, but 1V, or 1.1V is useless.

I'm currently torture testing a Mobile XP 2600+ that's been reduced to 1.6GHz @ 1.25V. If it passes, I'll reduce the voltage further and test some more. But I too, am limited by the 1.1V barrier.

PS - I used to be able to run it at 1.1V, but after a weird crash this past weekend, I'm not taking any chances without a 24 hour Prime95 test.

0Byte · Post by **0Byte** » Sat Aug 14, 2004 4:31 am

Well, I hope that a T-Bred B 1700+@500Mhz will run at 1.00V. This beast has already been proven to run 500Mhz @ 1.10V: http://www.silentpcreview.com/article164-page1.html
In any case I'm willing to give it a try. So is here anybody that knows a board that is able to set the Vcore to at least 1.00V or any kind of mod to do so?
It's really sad that most mainboard manufacturers support overvolting nowadays whereas undervolting seems to be still thought to be useless.

0Byte

mczak · Post by **mczak** » Sun Aug 15, 2004 3:48 pm

1.1V is some sort of a "natural limit", since the voltages defined by the VID signals range from 1.1V to 1.85V for Athlon XP cpus (they go down to less iirc on mobile cpus, too lazy to check). So typical voltage regulators on boards will only be able to provide voltages from 1.1V to 1.85V for simpler implementation (each 5-bit VID setting directly corresponds to a voltage).
It should be possible to "cheat" the voltage regulator by playing around with the feedback pin I think, there used to float around some mods for increasing the voltage above the bios provided range, the same methology might work for decreasing it.

peacefulpc · Post by **peacefulpc** » Sun Aug 15, 2004 4:29 pm

the nf7 was pretty good, with a minimum of 1.1 volts (my mobile athlon wasn't very stable below 1.25v though), but the nf7-s2 is useless. barring a bios revision, i think the lowest it goes is 1.5v, ug.

mr pink · Post by **mr pink** » Sun Aug 15, 2004 4:50 pm

surely not mine! I've got a msi kt4v and it only goes down to 1.625.

nmuntz · Post by **nmuntz** » Sun Aug 15, 2004 7:55 pm

My motherboard (cheapass soltek dual channel NF2) goes down to 1.1. I have a Athlon XP Mobile running at 800MHz @ 1.1v and its perfectly stable. I've had it converting mpeg2 to xvid for 36 hours straight without a hiccup. If the mobo could go down lower than 1.1, I would certainly try it.

Gxcad · Post by **Gxcad** » Sun Aug 15, 2004 11:56 pm

I have a Abit KD7 and it goes from 1.1 to 2.325v (great range!). I'd gladly give up all of the top end if I can run at or below 1v.

For the record I run my 1700+ tbred B at 1.16ghz at 1.1v for over a year now with absolute stability. It passes all torture tests with ease and can do the same at 1.2ghz. XP-M's should do even better!

-Ken

0Byte · Post by **0Byte** » Tue Aug 17, 2004 1:34 am

Thank you for your answers.
I didn't know about this VID code and have just asked because I've read twice about mainboards that were said to support a Vcore of 1.0 V:
Once about "all recent Soltek boards" which has turned out to be not true at least for all recent Socket A ATX boards, the other time about the K7SOM+, but sadly that's an µATX board and no information about the voltage range is available at the website of the manufacturer.

Gxcad wrote:For the record I run my 1700+ tbred B at 1.16ghz at 1.1v for over a year now with absolute stability. It passes all torture tests with ease and can do the same at 1.2ghz. XP-M's should do even better!

That's amazing! I haven't even dared to dream of such a result! Which stepping have you got?

0Byte

Extremely undervoltable Socket A board?

Extremely undervoltable Socket A board?

abit nf7 v. nf7-s2