The second success story :p

You can write down your success stories here (temperature decreasements, power savings, etc).
Post Reply
rubenvb
Posts: 8
Joined: Mon 6. Jul 2009, 16:04

The second success story :p

Post by rubenvb » Mon 6. Jul 2009, 16:20

I have just completed installing the phc-intel module on my Kubuntu Jaunty system (Pentium M 2.0 GHz).

In order for everything to work, I had to recompile the kernel as instructed here: http://www.linux-phc.org/forum/viewtopic.php?f=13&t=67

This worked the second time I tried it. I forgot to uncheck p4-clockmod and that got loaded instead of acpi_cpufreq. My stupid fault that forced me to rcompile the kernel under unbearably hot circumstances (100°C...).

My system is a HP nx8220 laptop. This is an old POS, but nonetheless good to get my feet wet in linux this summer.
The Pentium M ran at a default 1.365V for the 15x multiplier (2GHZ). Even in Windows this was unbearable and I used RMclock there to easily find the lowest stable voltage, which was..... 1.1V (!)
Thats almost a 20% difference. To document the temperatures, I needed to use the thermal zones the laptop has (there are four, which I think I have deciphered). I will report the two highest temperatures because I think they're CPU core and CPU ambient temp (the Tcore and Tcpu).
Before, at default 1.365V:
idle: Tcpu: 66°C, Tcore: 63°C
load (recompiling my Jaunty kernel): Tcpu: 100°C, Tcore: 94°C
These are abysmal and even dangerous, hence my undervolt:
After, at 1.1V undervolt
idle: Tcpu: 40°C, Tcore: 50°C
load (compiling that kernel again, for fun :p and this report): Tcpu: 55°C, Tcore: 65°C max

I don't have to illustrate any further, the result is quite clear... thanks for that abysmal cooling, HP
PS: fan speeds are down too, of course

v0lle85
Posts: 9
Joined: Fri 12. Jun 2009, 23:51

Re: The second success story :p

Post by v0lle85 » Wed 8. Jul 2009, 10:19

Wow that's a huge difference (and must be a huge difference in power consumtion, too)!

Did you also make a torture test, e.g. with mprime? Compiling is a good first test, of course, but I think you should make sure that something like mprime works correctly for some hours. On my Phenom system I sometimes got errors after 8 hours or so. ;)

Greetings,
Volker

rubenvb
Posts: 8
Joined: Mon 6. Jul 2009, 16:04

Re: The second success story :p

Post by rubenvb » Wed 8. Jul 2009, 23:04

Yeah, I probably should test more, but I don't feell like it. I suppose Kubuntu Karmic Alpha would crash before my cpu gets to making errors :p

(nevertheless, you are completely correct :) )

The figures speak for themselves, the cooling in this (stone age) laptop sucks very very.... very much.

tntprog
Posts: 16
Joined: Wed 7. Jan 2009, 02:23

Re: The second success story :p

Post by tntprog » Wed 15. Jul 2009, 14:26

My girlfriend has a HP laptop too, these computers are very crappy about CPU temperature! 98°C is "normal" for her laptop, even with RMClock (Windows).

Mrkva
Posts: 1
Joined: Thu 27. Aug 2009, 01:10

Re: The second success story :p

Post by Mrkva » Thu 27. Aug 2009, 01:11

tntprog wrote:My girlfriend has a HP laptop too, these computers are very crappy about CPU temperature! 98°C is "normal" for her laptop, even with RMClock (Windows).
No, that's not normal. Have you tried cleaning fan and cooling system?

Post Reply