|
|
|
|
|
|
| Author |
Message |
igor@nospam.invalid *nix forums addict
Joined: 17 Aug 2005
Posts: 81
|
Posted: Thu Jun 29, 2006 4:47 pm Post subject:
"cool running" geode processor
|
|
|
Hello.
An easy question, though! I have a Soekris Engineering net4801 running
OpenBSD 3.9. My goal is making that machine run as cool as possible to
avoid possible hardware damage to it. It is certainly not running very
hot:
$ sysctl hw.sensors
hw.sensors.0=nsclpcsio0, TSENS1, temp, 127.00 degC / 260.60 degF
hw.sensors.1=nsclpcsio0, TSENS2, temp, 127.00 degC / 260.60 degF
hw.sensors.2=nsclpcsio0, TNSC, temp, 56.00 degC / 132.80 degF
hw.sensors.3=nsclpcsio0, VSENS0, volts_dc, 3.01 V
hw.sensors.4=nsclpcsio0, VSENS1, volts_dc, 2.07 V
hw.sensors.5=nsclpcsio0, VSENS2, volts_dc, 2.55 V
hw.sensors.6=nsclpcsio0, VSENS3, volts_dc, 0.60 V
hw.sensors.7=nsclpcsio0, VSENS4, volts_dc, 2.54 V
hw.sensors.8=nsclpcsio0, VSENS5, volts_dc, 2.54 V
hw.sensors.9=nsclpcsio0, VSENS6, volts_dc, 0.00 V
hw.sensors.10=nsclpcsio0, VSB, volts_dc, 3.35 V
hw.sensors.11=nsclpcsio0, VDD, volts_dc, 3.35 V
hw.sensors.12=nsclpcsio0, VBAT, volts_dc, 3.01 V
hw.sensors.13=nsclpcsio0, AVDD, volts_dc, 3.33 V
hw.sensors.14=nsclpcsio0, TS1, volts_dc, 1.56 V
hw.sensors.15=nsclpcsio0, TS2, volts_dc, 1.59 V
hw.sensors.16=nsclpcsio0, TS3, volts_dc, 1.50 V
As its internal HDD (a 60GB Travelstar drive) is probably not designed
for 24/7 use, I think that it is good spinning it down when not in use:
$ tail -2 /etc/rc.local
atactl /dev/wd0c setidle 180; atactl /dev/wd0c setstandby 240
atactl /dev/wd0c writecachedisable
(writecachedisable is here to avoid filesystem corruption as this machine
has softdep enabled too... but I think that it is an off-topic on this
thread, just want to check that I did not a mistake enabling this option!)
My question is... makes it sense enabling "cool running" in apmd?
$ apm
Battery state: unknown, 0% remaining, unknown life estimate
A/C adapter state: not known
Performance state: cool running (267 MHz)
This machine has a geode processor. I haven't tried it on other
architectures but I would expect either a change in the processor
speed or a change in the hw.setperf variable---it seems that this
parameter is not supported, as it is not shown when using sysctl( .
Is cool running supported by AMD geode embedded computers?
Best regards,
Igor. |
|
| Back to top |
|
 |
mike *nix forums Guru Wannabe
Joined: 04 Mar 2005
Posts: 233
|
Posted: Sat Jul 01, 2006 8:38 am Post subject:
Re: "cool running" geode processor
|
|
|
Igor Sobrado wrote:
| Quote: | Hello.
An easy question, though! I have a Soekris Engineering net4801 running
OpenBSD 3.9. My goal is making that machine run as cool as possible to
avoid possible hardware damage to it. It is certainly not running very
hot:
$ sysctl hw.sensors
hw.sensors.0=nsclpcsio0, TSENS1, temp, 127.00 degC / 260.60 degF
hw.sensors.1=nsclpcsio0, TSENS2, temp, 127.00 degC / 260.60 degF
hw.sensors.2=nsclpcsio0, TNSC, temp, 56.00 degC / 132.80 degF
hw.sensors.3=nsclpcsio0, VSENS0, volts_dc, 3.01 V
|
I don't think you need to worry about running cool. At least as
far as cpu/memory damage is concerned.
From two of these beasties I have, one of which is running flat out,
and the other idle (backup) I see that the temp difference as follows.
idle - pretty much the same as yours.
$ sysctl hw.sensors
hw.sensors.0=nsclpcsio0, TSENS1, temp, 127.00 degC / 260.60 degF
hw.sensors.1=nsclpcsio0, TSENS2, temp, 127.00 degC / 260.60 degF
hw.sensors.2=nsclpcsio0, TNSC, temp, 60.00 degC / 140.00 degF
hw.sensors.3=nsclpcsio0, VSENS0, volts_dc, 3.01 V
busy
$ sysctl hw.sensors
hw.sensors.0=nsclpcsio0, TSENS1, temp, 127.00 degC / 260.60 degF
hw.sensors.1=nsclpcsio0, TSENS2, temp, 127.00 degC / 260.60 degF
hw.sensors.2=nsclpcsio0, TNSC, temp, 74.00 degC / 165.20 degF
I do not know what effect a hard disk drive will have as I haven't got
one.
I also don't don't know if it is possible to have a variable clock
frequency with apm. I can find no reference in the doc I have.
Anyway it messes up NTP, so I wouldn't use it.
Anyway it hasn't missed beat. |
|
| Back to top |
|
 |
igor@nospam.invalid *nix forums addict
Joined: 17 Aug 2005
Posts: 81
|
Posted: Sat Jul 01, 2006 12:34 pm Post subject:
Re: "cool running" geode processor
|
|
|
mike <michael.no.spam.cook@wanadoo.fr> wrote:
| Quote: | Igor Sobrado wrote:
$ sysctl hw.sensors
hw.sensors.0=nsclpcsio0, TSENS1, temp, 127.00 degC / 260.60 degF
hw.sensors.1=nsclpcsio0, TSENS2, temp, 127.00 degC / 260.60 degF
hw.sensors.2=nsclpcsio0, TNSC, temp, 56.00 degC / 132.80 degF
hw.sensors.3=nsclpcsio0, VSENS0, volts_dc, 3.01 V
I don't think you need to worry about running cool. At least as
far as cpu/memory damage is concerned.
|
Hi, Mike.
You are right. Power dissipation on a 5 Watt device is not a concern.
I was just trying to avoid too fast temperature variations when this
device changes from idle to busy. There are some physical concerns
related with temperature changes and surface mounting components
damage, and wanted to minimize it.
| Quote: | From two of these beasties I have, one of which is running flat out,
and the other idle (backup) I see that the temp difference as follows.
idle - pretty much the same as yours.
$ sysctl hw.sensors
hw.sensors.0=nsclpcsio0, TSENS1, temp, 127.00 degC / 260.60 degF
hw.sensors.1=nsclpcsio0, TSENS2, temp, 127.00 degC / 260.60 degF
hw.sensors.2=nsclpcsio0, TNSC, temp, 60.00 degC / 140.00 degF
hw.sensors.3=nsclpcsio0, VSENS0, volts_dc, 3.01 V
busy
$ sysctl hw.sensors
hw.sensors.0=nsclpcsio0, TSENS1, temp, 127.00 degC / 260.60 degF
hw.sensors.1=nsclpcsio0, TSENS2, temp, 127.00 degC / 260.60 degF
hw.sensors.2=nsclpcsio0, TNSC, temp, 74.00 degC / 165.20 degF
|
Same happens here. As running memtest86 on this machine is
certainly a challenging task, I compiled a large software package
on it. The temperature on that box increases in 9?C after ten
minutes.
| Quote: | I do not know what effect a hard disk drive will have as I haven't got
one.
|
I am using a fluid dynamic bearing based drive (a Hitachi Travelstar).
Don't expect a lot of problems related with these drives that run
really quiet and cold. Fluid dynamic bearing spindle motors are a
fine improvement for hard disk drives, IMHO. And these drives are
not expensive at all.
| Quote: | I also don't don't know if it is possible to have a variable clock
frequency with apm. I can find no reference in the doc I have.
Anyway it messes up NTP, so I wouldn't use it.
|
That is a very good point. I was asking about changing the clock
frequency using apm (or at least, introducing a lot of nop's when
load is high even if common sense shows that it seems a wrong thing
to do), but certainly a variable clock frequency will be a nightmare
for NTP. A lot of improvements in timedelta sensors dropped by
a change like this one! It is unnaceptable... :-)
| Quote: | Anyway it hasn't missed beat.
|
You really helped me a lot. I think that losing time synchronization
is something that must be considered. Perhaps it is better not running
apmd on that machine and stay at the default configuration.
Cheers,
Igor. |
|
| Back to top |
|
 |
Google
|
|
| Back to top |
|
 |
|
|
The time now is Tue Dec 02, 2008 1:15 pm | All times are GMT
|
|
Mobile Phone | Discount Magazines | Loan | eHarmony Coupon | Loans
|
|
Copyright © 2004-2005 DeniX Solutions SRL
|
|
|
|
Other DeniX Solutions sites:
Unix/Linux blog |
electronics forum |
medicine forum |
science forum |
|
|
Privacy Policy
|
Powered by phpBB © 2001, 2005 phpBB Group
|
|