[ltp] Fan always on problem... does it happen on R52 with the Intel Media Gr
Jaime Iniesta
linux-thinkpad@linux-thinkpad.org
Sun, 14 Aug 2005 14:57:08 +0200
------=_Part_2975_16845549.1124024228585
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline
Wouldn't it be better to monitor both temps? I mean:
if ((CPUtemp > criticalCPUtemp) or (GPUtemp > criticalGPUtemp)) then=20
{
start fan
}
Or, take an average measure based on both temps.
Jaime
2005/8/14, Paul RIVIER <paul.rivier@eleve.emn.fr>:
>=20
>=20
> >
> > This makes me wonder if the ibm_manage_fan script should be based on
> > GPU temperature, rather than CPU...? But I would have thought IBM
> > would base fan behaviour on the most critical thing that might melt,
> > which would be the CPU, rather than the graphics card, right?
>=20
> When I wrote this script, I wondered what temp I should monitor.
> Actualy, I found that GPU temp heavily depends on CPU temp. Seems to me
> that heat pipe is shared, or air flow go first through CPU, next through
> GPU. In both case, I get satisfying behavior using CPU temp only (I
> monitor all temperatures).
> Does anyone know any application that use GPU at 100% and no cpu at all
> ? That may be THE dangerous case.
> Thanks
>=20
> Paul
> --
> The linux-thinkpad mailing list home page is at:
> http://mailman.linux-thinkpad.org/mailman/listinfo/linux-thinkpad
>=20
--=20
Jaime Iniesta AKA Capit=E1n planeta
email: capitanplaneta ARROBA gmail PUNTO com
jabber: capitanplaneta ARROBA jabberes PUNTO org
webs: http://capitanplaneta.blogspot.com/
------=_Part_2975_16845549.1124024228585
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline
Wouldn't it be better to monitor both temps? I mean:<br>
<br>
if ((CPUtemp > criticalCPUtemp) or (GPUtemp > criticalGPUtemp)) then =
<br>
{<br>
start fan<br>
}<br>
<br>
Or, take an average measure based on both temps.<br>
<br>
Jaime<br>
<br><br><div><span class=3D"gmail_quote">2005/8/14, Paul RIVIER <<a href=
=3D"mailto:paul.rivier@eleve.emn.fr">paul.rivier@eleve.emn.fr</a>>:</spa=
n><blockquote class=3D"gmail_quote" style=3D"border-left: 1px solid rgb(204=
, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<br>><br>> This makes me wonder if the ibm_manage_fan script should b=
e based on<br>> GPU temperature, rather than CPU...? But I wo=
uld have thought IBM<br>> would base fan behaviour on the most critical =
thing that might melt,
<br>> which would be the CPU, rather than the graphics card, right?<br><=
br>When I wrote this script, I wondered what temp I should monitor.<br>Actu=
aly, I found that GPU temp heavily depends on CPU temp. Seems to me<br>
that heat pipe is shared, or air flow go first through CPU, next through<br=
>GPU. In both case, I get satisfying behavior using CPU temp only (I<br>mon=
itor all temperatures).<br>Does anyone know any application that use GPU at=
100% and no cpu at all
<br>? That may be THE dangerous case.<br>Thanks<br><br>Paul<br>--<br>The li=
nux-thinkpad mailing list home page is at:<br><a href=3D"http://mailman.lin=
ux-thinkpad.org/mailman/listinfo/linux-thinkpad">http://mailman.linux-think=
pad.org/mailman/listinfo/linux-thinkpad
</a><br></blockquote></div><br><br><br>-- <br><br>Jaime Iniesta AKA Capit=
=E1n planeta<br>email: capitanplaneta ARROBA gmail PUNTO com<br>jabber: cap=
itanplaneta ARROBA jabberes PUNTO org<br>webs: <a href=3D"http://capitanpla=
neta.blogspot.com/">
http://capitanplaneta.blogspot.com/</a>
------=_Part_2975_16845549.1124024228585--