Aspire One vs KillAWatt - Power Consumption Test (with pic)
- Thread starter vautrin44
- Start date
Wouldn't you also need the power drain of other netbooks to support this statement?
-John
-John
spyderms said:
If I were writing a paper on consumer notebook power consumption, yes.
If I claimed that this is the most energy efficient consumer notebook on the market, I should.
I assume most readers will accept a 9W idle power consumption for a consumer notebook a very good number.
Now, if you think a 9W idle power consumption for a consumer notebook is not a very low number and doesn't makes the AAO one of the most energy efficient notebooks out there, in a world of 80-100W desktop and 25-40W laptop idle power consumption, frankly, I don't understand what is your point. The ASUS EEE 701 4G, for example, idled at 14.5, a very good number (http://www.computerpoweruser.com/editor ... F13c01.asp), though not as low as the AAO.
If someone knows of a consumer notebook with a lower idle, please post. This is not a "AAO fanboy-oriented" post, just very happy at the great numbers my AAO did.
Under Linpus, the power consumption is displayed on screen when you click on the battery indicator. The numbers are in line with what you measure with your meter, suggesting 1) the onscreen data is accurate and 2) the power supply losses are minimal, say 20%.
Let's call the typical idle power consumption of the Aspire One 9W. Is this exceptionally low for a notebook? No.
My Thinkpad X31 idled at 8W. My T60 about 15W. (numbers corrected to compare with your meter readings)
Considering the small screen and considering the supposed power saving technology of the Atom CPU, it's a little disappointing that the X31 (Pentium M 1.3 Ghz, 12" cold cathode display, 20GB HDD, 512MB Ram) from 2003 consumes less power than Atom + SDD + 8.9" LED display from 2008.
If you are an optimist, you can think of it in a positive light: we can go to a 1.6Ghz CPU a much brighter display, and much more powerful graphics, and keep the consumption down to essentially the same level.
Still, I'd like to see idle consumption get down to about 5-6 watts, since that would mean we would see 4-5h use on the 3-cell battery and as a further benefit we could probably get rid of the fan...
Let's call the typical idle power consumption of the Aspire One 9W. Is this exceptionally low for a notebook? No.
My Thinkpad X31 idled at 8W. My T60 about 15W. (numbers corrected to compare with your meter readings)
Considering the small screen and considering the supposed power saving technology of the Atom CPU, it's a little disappointing that the X31 (Pentium M 1.3 Ghz, 12" cold cathode display, 20GB HDD, 512MB Ram) from 2003 consumes less power than Atom + SDD + 8.9" LED display from 2008.
If you are an optimist, you can think of it in a positive light: we can go to a 1.6Ghz CPU a much brighter display, and much more powerful graphics, and keep the consumption down to essentially the same level.
Still, I'd like to see idle consumption get down to about 5-6 watts, since that would mean we would see 4-5h use on the 3-cell battery and as a further benefit we could probably get rid of the fan...
I need to get one of those. I do know my eeepc sure puts out a lot of heat for its little size.
Once I hooked up a Fluke multimeter to measure the current from an old laptop. Turning on the caps lock LED showed up an entire 10mA. Every tiny thing it did showed up on the meter.
Once I hooked up a Fluke multimeter to measure the current from an old laptop. Turning on the caps lock LED showed up an entire 10mA. Every tiny thing it did showed up on the meter.
goofball said:
Boot but do not touch...
goofball said:
Mine runs idle most of the time. Load average is around 0.00. But I don't dare to suspend it...
Mine also happens to be my mailserver, DVR for the security webcams, and a monitor for the wireless network. If I have mine in suspend, mailing lists will start unsubscribing me, etc... Its always listening to connections and answering them.
This little laptop is x100 times more powerful than the server I had 10 years ago! Built in battery backup too!
in linpus linux you can install the tiny but useful powertop program
Run powertop as root
You only get power stats on battery power, so remove the ac power cable.
If I run 'glxgears' (from another terminal) usage goes up to only 15W, which is pretty impressive.
Unfortunately you don't get power stats per process since the linpus kernel is compiled without the CONFIG_TIMER_STATS option.
(Why they went to so much trouble to slim down the kernel and remove nearly all the kernel modules, and even remove support for ext3, and then added a ridiculously bloated openoffice installation, I have no idea. Note that ext2 takes ages to run a fsck on the 120gb hard disk on boot - you have to wait a few minutes with a blank screen after a system crash for the filesystem to be checked, an ext3 filesystem would recover almost instantly)
Code:
sudo yum -y update fedora-release
sudo yum install powertopRun powertop as root
Code:
sudo powertopYou only get power stats on battery power, so remove the ac power cable.
Code:
PowerTOP version 1.8 (C) 2007 Intel Corporation
Cn Avg residency P-states (frequencies)
C0 (cpu running) ( 7.8%) 1.60 Ghz 0.0%
C1 0.0ms ( 0.0%) 1333 Mhz 0.0%
C2 4.5ms (24.4%) 1066 Mhz 0.0%
C3 5.7ms (67.7%) 800 Mhz 100.0%
Wakeups-from-idle per second : 172.1 interval: 10.0s
Power usage (ACPI estimate): 10.7W (2.5 hours)
No detailed statistics available; please enable the CONFIG_TIMER_STATS kernel option
This option is located in the Kernel Debugging section of menuconfig
(which is CONFIG_DEBUG_KERNEL=y in the config file)
Note: this is only available in 2.6.21 and later kernels
Suggestion: Enable the CONFIG_HPET_TIMER kernel configuration option.
Without HPET support the kernel needs to wake up every 20 milliseconds for
some housekeeping tasks.
Q - Quit R - RefreshIf I run 'glxgears' (from another terminal) usage goes up to only 15W, which is pretty impressive.
Unfortunately you don't get power stats per process since the linpus kernel is compiled without the CONFIG_TIMER_STATS option.
(Why they went to so much trouble to slim down the kernel and remove nearly all the kernel modules, and even remove support for ext3, and then added a ridiculously bloated openoffice installation, I have no idea. Note that ext2 takes ages to run a fsck on the 120gb hard disk on boot - you have to wait a few minutes with a blank screen after a system crash for the filesystem to be checked, an ext3 filesystem would recover almost instantly)
My AAO was working at about 20W while playing a 15MB MP4 video file. I'll post the pic later.
That number doesn't impress me much. I expected 15-17W. Still, my TravelMate idles at about 30W.
I was checking out the OLPC specs and, apparently, it idled at 1W.
I'll do more testing with my AAO (OpenOffice, browsing, youtube, etc.) later. Of course, I'll post pics.
MJH - I didn't knew the X-31 was that good at energy efficiency. Great numbers indeed.
melhiore - I am not up to date with the Windows AAO, but your machine sounds great. I am very interested in it since my significant will buy a Windows AAO soon.
That number doesn't impress me much. I expected 15-17W. Still, my TravelMate idles at about 30W.
I was checking out the OLPC specs and, apparently, it idled at 1W.
I'll do more testing with my AAO (OpenOffice, browsing, youtube, etc.) later. Of course, I'll post pics.
MJH - I didn't knew the X-31 was that good at energy efficiency. Great numbers indeed.
melhiore - I am not up to date with the Windows AAO, but your machine sounds great. I am very interested in it since my significant will buy a Windows AAO soon.
melhiore said:goofball said:
Boot but do not touch...
Do not touch or do not run anything? There's a difference between the two, isn't there? What I have running is probably different than what other's are running as default processes/services. Running it as a server, it's not ever truly "idle" since it's always doing "something" (listening)?
I guess technically the OS is always doing "something" though... :?:
I got the Kill A Watt device yesterday. Found out the power factor was about 50%. That means the power supply is clipping the AC waveform and causing a reflection back through the power meter out back, making me pay almost twice as much for the electricity. I'm going to size up a capacitor and improve the power factor.
dattaway said:
Your electricity provider charged you based on your power factor? That sounds unusual for a residential supply, it's unheard of in the UK at least. You could add power factor correction to the Acer power supply, but it would be neater and possibly cheaper to simply replace it with a 19v supply that has a PF that you like.
I'd suggest checking with the supplier first and see if it's worth it, I mean 9watts being double charged (if you are), lets say that its plugged in and running 24 hours per day, all year, that's 24x365 = 8760hours. What's the price per KHw where you are? Let's say it's 10 cents. The extra 9w that you may be getting charged for would then cost (8760x9) / 1000 = 78.84 KWh. At 10cents each one, that's gonna run you almost 8 bucks a year. If you never unplug it.
How much with PFC cost?
My AAO with bluetooth module and SD card inserted, pulls 20-21 watts idle. Without the BT module, it's 17 watts.
Under a full CPU load (Prime95) it's only drawing 23 watts. Playing a DVD VOB file with the backlight turned up and with sound pulls 25 watts.
Under a full CPU load (Prime95) it's only drawing 23 watts. Playing a DVD VOB file with the backlight turned up and with sound pulls 25 watts.
