[sf-lug] will increasing RAM increase power draw during suspend-to-RAM?

[Mark K. Zanfardino]

There have been some really good and interesting comments in the thread 
concerning this issue.  I should state that my laptop is a ThinkPad T42 
recently (in the last year) refurbished by my wife's company before 
being made available to be purchased by the employees following a 
massive hardware upgrade.  Nice deal too, only cost $160!

Anyway, I might suspect the age of the battery, except that the time 
differential between discharge while in sleep mode has easily doubled 
following the aforementioned upgrades.

The explanation of the battery draw as a function of current and 
resistance helps to shed some light on the subject.  As I recall from 
"What every programmer should know about memory" 
(http://lwn.net/Articles/250967/), chapter 2.12 - Dynamic RAM - gives a 
clear explanation as to how modern DRAM works and the need for it to be 
frequently "refreshed" (note that the DRAM cell design is centered 
around a single transistor and a single capacitor which requires 
refreshing to maintain state).  As a result, power will be consumed.  In 
my case I have added 200% RAM (1GB in addition to the 512MB already 
on-board).  I suppose it stands to reason that the amount of power 
required to refresh the RAM in my instance should have tripled - though 
it may not be a linear relationship.

I like the ideal proposed by John about simply not putting power to the 
additional RAM until needed.  However, I would think that would have to 
be something controllable at the BIOS level.  The kernel will identify 
the amount of RAM available at boot time (I assume) based on what the 
BIOS manages.  Unless there is a hook that could permit this block of 
RAM to be "turned off", I don't see how it could be made dynamic.  But 
it's a neat idea.

Mark

Alden Meneses wrote:
> Could it be other factors like the age of the battery? I have not 
> changed anything on my laptop in regards to the hardware and the 
> battery life is not what it used to be.
>
> On 9/4/08, *Mark K. Zanfardino* <mzanfardino at gmail.com 
> <mailto:mzanfardino at gmail.com>> wrote:
>
>     John,
>
>     This is a very good question and it brought to mind the fact that my
>     laptop does not maintain it's battery life nearly as long as it had
>     prior to my adding an addition 1GB SODIMM to the existing 512MB
>     base RAM
>     installed.  Prior to the expansion I could safely leave my laptop in
>     sleep mode for nearly 24 hours; now it doesn't last 12.  I haven't run
>     any conclusive tests to see how long it will last, but clearly
>     there has
>     been a dramatic decrease.
>
>     NOTE: I wasn't looking at the RAM initial for this decrease as
>     I've made
>     a number of changes including installing a larger HD (moved from a
>     40GB
>     4400rpm drive to a 120GB 5400rpm drive) and installing a newer release
>     of Kubuntu (I was running 7.04 and am now running 8.04.1).  It seems
>     unlikely that the drive would be the issue, as it should be parked
>     while
>     sleeping.  The OS might be an issue, I guess, but I can't be sure.
>     However it seems the RAM the likely suspect.
>
>     I'd be interested in reading any other's thoughts on this issue.
>
>     Mark
>
>     John Magolske wrote:
>     > When suspended to RAM, a laptop's battery is slowly drained.
>     > My understanding is that this power is going towards keeping
>     > the RAM active. I'm wondering how much of the total power draw
>     > is is being consumed by the RAM chips. Would doubling the
>     > amount of RAM from say 512MB to 1GB have the effect of roughly
>     > doubling the power consumption while suspended?
>     >
>     > John
>     >
>     >
>     >
>
>     _______________________________________________
>     sf-lug mailing list
>     sf-lug at linuxmafia.com <mailto:sf-lug at linuxmafia.com>
>     http://linuxmafia.com/mailman/listinfo/sf-lug
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> sf-lug mailing list
> sf-lug at linuxmafia.com
> http://linuxmafia.com/mailman/listinfo/sf-lug
>