Atom schmatom; folks with a KillaWatt, how low can you get your system idle and load?

[pm]

Do you have an active PFC power supply, IDC? Those are the type that Zap said fool the Kill-o-watt.

Personally, I like to think that I have a decent understanding of active power and a general understanding of how a kill-o-watt works, and I don't see how an APFC power supply would "fool" a kill-o-watt into wildly inaccurate readings. The videos are pulled from YouTube, and I didn't sign up at Johnyguru to view the photos, but the final photo isn't that far off. 219W on the SM-8800, 215W on the Kill-o-watt. I can live with a 2%... it could just be the way they are averaging over time.

 

[Roy2001]

My MSI Atom barebone system with 750GB HD, which is being used as a server and downloading 24/7 (to USB flash drive), idled at 19W.

 

[alkalinetaupehat]

On Topic:
Rig in the Sig gets 115w idle (per KAW) and about 330w loaded (FurMark and Prime95 64-bit). Not bad for a machine which gets 85fps in L4D (Max settings, 1680x1050) and does ~500ns/day on F@H.

Off Topic:
I skimmed the JonnyGURU thread and from what I got, the issue was that power supplies with Active PFCs were supposedly getting ~100% efficiency, it seems the Active Power Fluctuation Correctors in modern PSUs are causing the Kill-A-Watt to read the draw from the PC not accounting for PSU inefficiencies by doing something to the sine waves in the line which cancels out the additional load from the PSU's inefficiencies.

This whole issue about accuracy is ironic because if the APFCs are actually causing KAWs to measure straight component draw and effectively remove power supply efficiency from the energy draw equation, then hobbyists now have a $20 tool which can "accurately" measure the actual change from component changes and with careful work, enthusiasts can determine the actual power draw of individual components within a PC.

The irony of course comes from some enthusiasts complaining about measurements factoring in PSU efficiency and calling for measuring a component's individual power draw, while other enthusiasts want a tool which will measure total power draw from all PC components.

 

[LOUISSSSS]

lets talk about the ATOM

can someone with any netbook post your numbers of idle/load with the following information:
1. specs of the netbook
2. power draw from wall a few minutes of idling after boot up
3. power draw from wall while browsing internet/downloading/listening to music/watching video (tell us which)

thank you!

 

[mooseracing]

Originally posted by: dug777
Load for gaming was about 115W (1.1vcore and EPU-6 running).

Saw some numbers for Atom machines that didn't seem THAT much better for idle, given how piss-weak it is.

Suspect someone with a 45nm C2D or an equivalent AMD chip and an IGP will win this one easily

Now I've replaced the 3450 with a 4850 my idle and light use (office/net/email/winamp) is around 110W, and my gaming load is around 230W

Of course, I can now play games

You number sure seem low.

Our Dell Opti 755's with C2D @ 3Ghz, 4 sticks of ram, 2 HD's, and a Ati 2400 card draw about 80-90 at idle with 2 LCd's hooked up as well. At load they push about 140-150. Our P4 towers idle at about 100 watts.

My P-M 1.4 laptop will idle at 6-12 watts.

I'm not sure about by dual Opteron rig, as it is down at the moment.

 

[Idontcare]

Originally posted by: pm
Do you have an active PFC power supply, IDC? Those are the type that Zap said fool the Kill-o-watt.

Personally, I like to think that I have a decent understanding of active power and a general understanding of how a kill-o-watt works, and I don't see how an APFC power supply would "fool" a kill-o-watt into wildly inaccurate readings. The videos are pulled from YouTube, and I didn't sign up at Johnyguru to view the photos, but the final photo isn't that far off. 219W on the SM-8800, 215W on the Kill-o-watt. I can live with a 2%... it could just be the way they are averaging over time.

Actually I do, and you know for the life of me I can't recall whether I plugged the KWA into the circuit upstream or downstream my active APC unit...

When the results matched the utility company (I love them, they log power consumption by the hour on my webaccount so I can easily isolate power consumption by running simple tests at night when the background load is essentially unchanging) I rapidly jumped to the conclusion portion of my studies Could be I was fooling myself the whole time, I've been wronger than wrong before.

 

[pm]

Originally posted by: LOUISSSSS
lets talk about the ATOM

can someone with any netbook post your numbers of idle/load with the following information:
1. specs of the netbook
2. power draw from wall a few minutes of idling after boot up
3. power draw from wall while browsing internet/downloading/listening to music/watching video (tell us which)

thank you!

Using an Acer Aspire One (Atom N270 1.6GHz, 9" LCD, Intel 945GSE chipset, 512MB, 8GB SSD - this one: http://www.newegg.com/Product/...x?Item=N82E16834115530 )

I read:
11W in BIOS
15W booting (peak)
13W idling about 2 minutes after boot
13-14W browsing the web while playing an MP3 (it bounced back and forth, but mostly was at 13W)

By the way, to avoid confusion, when I mentioned that my Atom motherboard was reading 54W, it was this board ( http://www.newegg.com/Product/...x?Item=N82E16813121359 ) not an Atom Netbook. I have both - I bought the Atom to use as a HTPC/file server (it's waiting to be Ebay'd - I switched to the Pentium M system that I mentioned above), and then I bought the Netbook for my daughter to use for her schoolwork.

Originally posted by: Idontcare
Actually I do, and you know for the life of me I can't recall whether I plugged the KWA into the circuit upstream or downstream my active APC unit...

When the results matched the utility company (I love them, they log power consumption by the hour on my webaccount so I can easily isolate power consumption by running simple tests at night when the background load is essentially unchanging) I rapidly jumped to the conclusion portion of my studies Could be I was fooling myself the whole time, I've been wronger than wrong before.

I'm just asking to try to figure out if it's in apples to apples with what Zap was posting. If your kill-o-watt matches the company meter(that's really cool that they do that by the way... wish mine did), but your power supply isn't active PFC then it's not the same as the case that he was describing. What's an APC? You have an active power factor correction gizmo that can be plugged in in front of the power supply? I thought that active PFC was a capability of the power supply going from AC -> DC.

 

[Idontcare]

Originally posted by: pm
What's an APC? You have an active power factor correction gizmo that can be plugged in in front of the power supply? I thought that active PFC was a capability of the power supply going from AC -> DC.

No no, nothing like that. This is a case of me being wronger than wrong.

I completely misinterpreted what you were talking about so my post pretty much is utter nonsense. (opened mouth, proved Lincoln's quip continues to stand the test of time)

I saw the word "active" and assumed we were talking power-backup with inline double-conversion line conditioner...which was about as far off the mark as it gets...soooo I'll extract myself from this hole I have dug and fade back into being a reader of the thread instead of a misinformer .

 

[Lord Banshee]

Originally posted by: pm

Originally posted by: LOUISSSSS
lets talk about the ATOM

can someone with any netbook post your numbers of idle/load with the following information:
1. specs of the netbook
2. power draw from wall a few minutes of idling after boot up
3. power draw from wall while browsing internet/downloading/listening to music/watching video (tell us which)

thank you!

Using an Acer Aspire One (Atom N270 1.6GHz, 9" LCD, Intel 945GSE chipset, 512MB, 8GB SSD - this one: http://www.newegg.com/Product/...x?Item=N82E16834115530 )

I read:
11W in BIOS
15W booting (peak)
13W idling about 2 minutes after boot
13-14W browsing the web while playing an MP3 (it bounced back and forth, but mostly was at 13W)

By the way, to avoid confusion, when I mentioned that my Atom motherboard was reading 54W, it was this board ( http://www.newegg.com/Product/...x?Item=N82E16813121359 ) not an Atom Netbook. I have both - I bought the Atom to use as a HTPC/file server (it's waiting to be Ebay'd - I switched to the Pentium M system that I mentioned above), and then I bought the Netbook for my daughter to use for her schoolwork.

Originally posted by: Idontcare
Actually I do, and you know for the life of me I can't recall whether I plugged the KWA into the circuit upstream or downstream my active APC unit...

When the results matched the utility company (I love them, they log power consumption by the hour on my webaccount so I can easily isolate power consumption by running simple tests at night when the background load is essentially unchanging) I rapidly jumped to the conclusion portion of my studies Could be I was fooling myself the whole time, I've been wronger than wrong before.

I'm just asking to try to figure out if it's in apples to apples with what Zap was posting. If your kill-o-watt matches the company meter(that's really cool that they do that by the way... wish mine did), but your power supply isn't active PFC then it's not the same as the case that he was describing. What's an APC? You have an active power factor correction gizmo that can be plugged in in front of the power supply? I thought that active PFC was a capability of the power supply going from AC -> DC.

PM,

is there a way to turn off the lcd and see how much it is taking up?

 

[pm]

Originally posted by: Lord Banshee
is there a way to turn off the lcd and see how much it is taking up?

I turned off the LCD (fn-F6) while it was idling in Windows and the power dropped by 2W down to 11W.

 

[dug777]

Originally posted by: mooseracing

Originally posted by: dug777
Load for gaming was about 115W (1.1vcore and EPU-6 running).

Saw some numbers for Atom machines that didn't seem THAT much better for idle, given how piss-weak it is.

Suspect someone with a 45nm C2D or an equivalent AMD chip and an IGP will win this one easily

Now I've replaced the 3450 with a 4850 my idle and light use (office/net/email/winamp) is around 110W, and my gaming load is around 230W

Of course, I can now play games

You number sure seem low.

Our Dell Opti 755's with C2D @ 3Ghz, 4 sticks of ram, 2 HD's, and a Ati 2400 card draw about 80-90 at idle with 2 LCd's hooked up as well. At load they push about 140-150. Our P4 towers idle at about 100 watts.

My P-M 1.4 laptop will idle at 6-12 watts.

I'm not sure about by dual Opteron rig, as it is down at the moment.

Are the LCDs/one of the LCDs powered through the PSU? I assume not.

I would hazard a guess that they may not be speedstepping?

My P4 tower at work idled around 70W with intel IGP and one HDD. Load iirc was around 130W.

Remember that my machine only has one HDD, and speedstepping/Asus EPU-6 hax drop the vcore to about .98 in idle/light use. My gaming load is around 230W (1.1875vcore BIOS, 1.12vcore CPU-Z).

EDIT:

GF's E8500/4GB/Gigabyte EP45 DS3L (P43)/9600GT/WD 640GB/PCI HDTV tuner/USB wireless idles at 92W.

It only drops to 2Ghz (higher FSB), and only speedsteps to 1.12v, so the 45nm advantage is pretty much nixed, difference will be the graphics cards.

 

[edplayer]

Tom's article on performance and efficiency

Using an E7200, a Foxconn G31MG-S and a Fortron FSP220 power supply they were able to get 31W measured at the wall for idle and a peak reading of 61W (using SYSMark 2007 to load the computer)

Could have dropped it some more with single channel memory and a different hard drive


Undervolting/clocking the cpu would have dropped it a little more