Power Usage: Lies & Statistics.

[nickpoore]

I've been reading more and more about people who look up the specs of the laptop and are making "informed" options based on these statistics.

Now, there are lies, [censored] lies, and statistics.

The stastics are meaningless.

To look at the basic specs.
The 8-cell battery has 75Wh of charge.
The Intel T2400 uses 31W of power.
Therefore, the laptop can only have about 2.5 hours of usage (assuming that the screen, hard drive & wireless tools use no power whatsoever.)

Clearly something is amis.

Lucily, Lenovo include a great Battery Meter tool.

I ran some tests.

With Wireless ON, Disk spinning, but not active, Processor ~5%, Screen brightness at MIN I use 14.8V, 0.89A (about 13W total)

With Wireless ON, Disk trashing, Processors maxed at 100%, screen brightnexx at MAX, I peaked at 14.8V, 1.45A (about 22W)**

With wireless OFF, Disk spinning not active, Processor <5%, Screen at MIN, the lowest numbers I have seen are 14.8V, 0.76A (about 11.2W).

So, the maximum battery life I can get will be about 6 hours 40 minutes.
The mimimum battery life I can get will be about 3 hours, 25 minutes.

The interesting thing to note here is that even when I was using everything I could, I had a tough time peaking above 22W and nowhere near to the 31W the processor was supposed to use.

I find that a typically, I get 4 to 6 hours of battery usage.

The best way I find to extend my battery charge is to proactively turn the screen brightness up and down based on ambient light conditions.

*Lastly - to explain my "maximum usage" test.
I have a program that processes digital images. I take RAW images with my Canon 1Dmk2 camera that are 8MB each.
My software (Phase One) reads the 8MB file, process it (about 20sec of 100% dual-processor utilization) and then saves a 5MB JPEG file to disk. This program uses lots of memory, processor and disk useage.

Hopefully, these "real world" numbers will help to explain the battery useage better than those numbers the marketing guys are throwing around.

-=Nick=-

[jimmcclarty]

That is really interesting, but I am curious about how the total power could be less than the intel published power usage for just the cpu. I wonder if the speed step took you machine's power usage down while on battery? Is there any way to know for sure?

[nickpoore]

Great - you found the flaw in my testing.

I had been running my testing in "power source optimized mode".

I re-ran the tests in Performance mode, and got the following results.

With Wireless ON, Disk trashing, Processors maxed at 100%, screen brightness at MAX, I peaked at 16.2V, 2.05A (about 33.2W)

Horrah, I finally got above 31W...

For those of you who would like to see a screenshot of my system during the testing, you can find it here:
http://www.nicholaspoore.com/x60screenshot.jpg

Oh, FYI, my X60 has a T2400, 1GB of RAM, 100GB 5400rpm HDD.
I have the windows swapfile turned off.

-=Nick=-

[CHoPSTICK89]

thanks for the intensive testing. Whats the battery life approx. when its at 31W?

[nickpoore]

thanks for the intensive testing. Whats the battery life approx. when its at 31W?

http://www.nicholaspoore.com/x60screenshot.jpg

[lithium726]

just so you guys know, the 31w intel states is for the entire processor family, so, if youre using a 2.13ghz Core Duo under something like Prime95 (ie, something that really maxes out both cores and cache), youll hit high 20's in wattage. 31w is for power viruses, or a program that will stress the CPU more than it could ever be stressed by any legit code. as an example, the 2.26ghz Pentium M (780) is a 27w processor. under prime, it consumes roughly 22w.

therefore, a 1.83ghz core duo running real world apps will consume considerably less power than 31w.