Undervolting G73jh HD5870M: Tell your observations

Alright. It's obvious that the HD5870M voltage settings are not hardware locked. At least it isn't for MSI as there are users who have noticed significant decrease in heat and have close to 3 hour battery life, with voltage as low as .70v. And since Alienware and Clevo are also running at 1.05v, suggest theirs isn't locked either, since all 3 ATi fabricated boards. Maybe that is the key here, Asus created an inferior board.

MSI users having undervolt success http://forum.notebookreview.com/msi/524841-undervolting-hd5870m-battery.html

The GPU Observer Gadget can read the voltage on your HD5870M. I know there is one user here at Asus forum who believes HD5870M for G73jh is hardware locked. But there are others who have observed the voltage changing.

Another user tried to change to .80v. The fact that his GPU kept locking up at boot up suggest that the HD5870M in fact did downclock to .80v but was unstable. http://forum.notebookreview.com/6794681-post23.html

So I think for those brave enough and those who want better GPU temp at idle and better battery life, post here of your results.

 

OK, fine, but what utility would you use to test voltages, other than RBE'ing?

 

No idea. The MSI crew use RBE.

I did use Nibitor and check out the GTX 285M which has similar performance as HD5870M, couldn't find the 460M vBios. But the GTX 285M run at 1.03v for 3D and .85v for 2D. And run at 100/150 for 2D/Battery. The voltage and frequency for 2D state for Asus HD5870M is WAY WAY too high, that's for sure. Question is, the vBios still broken from Asus or can the beta vBios can handle the voltage/frequencies like MSI, Clevo and Alienware can? Also have to ask, what did Asus do different when fabricating the boards for the HD5870M that ATi didn't do to mess it all up if it isn't a vBios issue?

I understand the fear here, considering how unstable and well broken this vBios appears to be. But if there is a brave soul, be awesome to some positive results. Be great to see negative results also as Asus can then understand, something is wrong. Personally wishing Gary and his Asus crew could be doing these testing themselves instead of relying on NBR community to troubleshoot these issues.