Hiya Rudolf (and group), OK, so if the AMD temp is incorrect, is there a software fix? I know AMD's Overdrive software can read the sensors (MS-Windows XP), but are they compensating for the inaccuracy somehow? Or just showing the inaccurate value? Is it low, high, or just plain wrong? ;-) If the temp sensors are wrong then how does the chip know to slow down at 70 C? ;-) My motherboard is an ASUS M3A78-T which I believe has temperature sensors on it. Would those temp sensors be supported by lm-sensors? (They're not being detected now). On the MS-Windows side, I think either Speedfan or Coretemp can read the motherboard sensors (besides ASUS proprietary software.) I'm sure it's not as simple as incorporating someone else's code into this application, but it seems it's been done and if we can avoid re-inventing the wheel it might be a benefit. Thanks in advance. Mike Doerner > ---------------------------------------------------------------------- > > Message: 1 > Date: Sat, 11 Apr 2009 13:25:27 +0200 > From: Rudolf Marek <r.marek at assembler.cz> > Subject: Re: AMD 10h core temperatures > To: Carsten Aulbert <carsten.aulbert at aei.mpg.de> > Cc: lm-sensors at lm-sensors.org > Message-ID: <49E07E27.30802 at assembler.cz> > Content-Type: text/plain; charset=UTF-8; format=flowed > > Hi, > > Unless I made some mistake in the overview the temps looks like very similar to > reported IPMI ones right? > > So far all fam10h CPUs have same bug making the temps via the k8temp driver > inaccurate. It might work for you it might not. This is the reason why is still > there no driver. > > Rudolf > >
