What's your computer's idle power consumption?

[APWNH]

I was pleasantly surprised to see the power draw at the wall on my newly upgraded system.

Using a Kill-A-Watt P3 (claims 0.2% accuracy -- I definitely have no way to verify that)

Specs in sig.

Stays right around 49W on the Windows 7 desktop.

This is only about three times the power consumed by my already very efficient CFL desk lamp. This idle power efficiency is really, really impressive given what the machine can do (just scored 7504 on 3DMark 11 -- next up of course is metering power during load).

3DMark11 Power consumption numbers are in: 259W peak during the final physics+graphics test. Looks to be a reasonable gaming load. During the first few graphics tests it only ever went up to 202W during the first benchmark.

[NeelyCam]

I have two - one is idling at 22W (i5-670), the other at 23W (i7-2600k). Mine are running at stock, though, using IGPs. Yours is hot stuff

[Jason181]

~230w idle. Ouch, eh? 2x6970s I think do it. 4x750GB Caviar Blacks, 2600k @ 4.5 Ghz.

[kyboshed]

Mac mini (2011 i5-2.3, 8GB), 11W idle, 16W playing local 1080p h264 video, 35W exporting the same video at 720p. Wifi was enabled throughout. <1W in standby (meter just reads 0W).

[rcs2k4]

I want to cry.

Idles at 125-131W, under load it fluctuate between 470-510W (was running Prime95, HDBench and Furmark all at the same time - Should be about as worst case as you can get).

Spec is in sig, and my power company love me [/sarcasm]...

Edit: Typos

[just brew it!]

I wonder how much of that is the FX and how much is the dual GPUs. Looks like it idles at about the same power level as my old dual Socket A Athlon MP rig (and power usage was the main reason I decided to retire that rig instead of using it as a file server).

[APWNH]

More numbers: 34W, idle with screen off. 238W, Furmark only.

[grantmeaname]

12W (best guess?). I win the internets!

[anotherengineer]

Well, mine is about 170W, however that includes everything that is plugged into the power bar, so
dual monitors both ccfl bulbs, one is 120hz,
all in one printer,
speakers,
headphones,
cable modem,
telephone, and
PC (amd 955 - stock, 6850, dedicated sound card, and network card, dvd burner, and 256gb plextor m3, oh and psu is fairly old, I think it's only 80+ (not even bronze)

bios - C1E is diabled, cnq disabled, and win7 x64, power settings to high performance mode.

With everything shut down it reads 25-32W being drawn lol

Each monitor drops it about 35W when going to sleep so that brings it to 100W, I guess if I plugged only the PC it would be around 70W, idle on desktop, then if you take off the efficiency loss from the psu, even less?

[Sargent Duck]

I don't know, but it's pretty low.

AMD E-350, 600GB Blue WD, 600GB Green WD, 60GB Crucial Force 3

[rcs2k4]

I wonder how much of that is the FX and how much is the dual GPUs. Looks like it idles at about the same power level as my old dual Socket A Athlon MP rig (and power usage was the main reason I decided to retire that rig instead of using it as a file server).

I think it's mostly the CPU too, as for whatever reason when it's overclocked, it won't clock down when idle. Put it back to stock speeds and it will clock down when idle, but then it's slow - Oh well. I was curious and tested out this discovery on power consumption:

FX CPU at Stock Clock (Everything else in system as it was before):
Idle: 81-84W
Load: 384-395W

Colour me a little shocked - Thats a lot different, but also a lot slower. The extra 1.6Ghz overclock is directly responsible for an additional average power consumption at idle of 40W, and under load an eye watering 105W. So thats the direct consiquence of overclocking an FX - It's power hungry. What I dont understand is why the power consumption difference at idle and load for just the CPU are so different. I would have thought if it was running 4.7Ghz idle, and same clock speed at load the power consumption of the chip would be the same?

[just brew it!]

Modern CPUs don't just ramp clock speed up and down in response to load, they also selectively shut down sections of the chip (entire cores, even!) when the load is light.

[DPete27]

do you have AMD Cool 'n' Quiet turned off rcs? Without cool 'n' quiet on, I don't believe the CPU will throttle its clocks down when idle, but what do I know, I'm an Intel guy. No Kill-a-Watt for me. I just can't justify the $20 for something so trivial...."oooh look, ____ watts, niice" It is what it is.

[just brew it!]

do you have AMD Cool 'n' Quiet turned off rcs? Without cool 'n' quiet on, I don't believe the CPU will throttle its clocks down when idle, but what do I know, I'm an Intel guy. No Kill-a-Watt for me. I just can't justify the $20 for something so trivial...."oooh look, ____ watts, niice" It is what it is.

If you live somewhere with expensive electricity I could see it paying for itself over the course of a year or so, if you use it to make intelligent decisions about your power usage.

[rcs2k4]

Cool 'n' Quiet is enabled, but only seems to kick in if you leave the CPU at stock. The slightest change to multiplier or bus speed and it doesnt seem to clock back the chip once idle... I'm not so fussed about the power consumption, yet - My PC works and does all that I want it to and my electricity bill is about what it has always been each month (had this CPU for 2 months now). It don't even get that hot, or loud. The difference in power draw when you see the figures is a bit shocking at how much it does vary, but I guess nothing new has been learnt here.

Who would have thought that the FX-8120 overclocked would consume a lot of power?

[indeego]

Home: ~190W including two 23/24" LCDs. i7-980X + 3 SSD/3 Mechanicals, a GTX 460.

Work: 11W, 2510p laptop Core2 L9400 @ 1.86GHz and a Droid Charge being Charged.

Second one seems low and wrong, but that's what my UPS says.

[chµck]

10-20w is my best guess. Using a 17" MSI GT735 with an extra monitor. External sound card, USB hub powering external keyboard and mouse. No more than ~120w under load.

[Kaleid]

66w with monitor on.

[APWNH]

I realized my macbook is able to sense the power consumption from its battery. It runs slightly under 7 watts idle with a low-ish backlight level: I could set it to the same settings and plug it into the power meter.

Unfortunately that would also measure the A/C adapter's wasted power so I'm not gonna be able to use this to check if my Kill-A-Watt is telling the truth. Not without a calorimeter or something.

[axeman]

Where does one obtain said kill-a-watt? I have a pretty high electricity bill. I'm in Canada, and my electricity is something on the order of $0.10/kwh. Most of our electrical generation here is coal-fired power stations. I miss Manitoba and their Hydro for 6.7 cents/kwh.

[APWNH]

Does Amazon sell in Canada? It's where I got mine for about $20.

10 cents is pretty good...

[axeman]

It looks like newegg.ca sells them. 10cents might not be bad compared to some places, but it does seem high compared to most of our neighboring provinces or states. I also have some pretty bad habits as far as leaving everything all the time. It would gain no small amount of points with the missus to get that bill down a little (120-150/mo for a 3 bedroom bungalow is pretty considering the hot water heater and furnace are gas ), and especially if I knew what devices would be especially beneficial to have switched off.

[rado992]

Around 180W for these components: Q9550 @ 3.83, X48 mobo, 4 sticks of RAM, 8800GTS 512, 5 HDDs, Xonar DX and a bunch of fans. Reading this topic makes me wonder if I should upgrade for the power savings alone.

[APWNH]

I do also run a i5-750 box with no less than 7 HDD's in it (the bays are filled). I reckon that one sucks a bit more juice.

[Triple Zero]

~100-110 W at idle, PC only (display not included).

[anotherengineer]

Where does one obtain said kill-a-watt? I have a pretty high electricity bill. I'm in Canada, and my electricity is something on the order of $0.10/kwh. Most of our electrical generation here is coal-fired power stations. I miss Manitoba and their Hydro for 6.7 cents/kwh.

I hear ya. In Ontario after, delivery, debt repayment fee, 13% HST electricity works out to about 17 cents/kWhr, about double Manitoba, yes feel the butt hurt. My mom had forced air electric heat, my brother and I got a nat. gas furnace to replace it about 8 months ago and it payed for itself in 3 months lol

I put it on an older fridge for fun and ended up at about $23/month (using 16cents/kwhr), I am in partial disbelief when I get the hydro bill every month and it says the family consumed over a megawatt-hr of power!!!!

I use this to check stuff out, it more accurate for resistive loads, the switching PSU's will make it bounce around occasionally.
http://www.canadiantire.ca/AST/browse/G ... ?locale=en
Watch for sales, I picked it up for $15 bucks.

[notfred]

In Ottawa the public library lends out Kill-A-Watt meters http://ottawa.bibliocommons.com/item/sh ... ill_a_watt
Check your library to see if they do the same thing

[Waco]

190 watts, but that's not including my 42" large format display (which pulls about 150 watts).

My machine is only on when I'm using it nowadays but it's still better than my 4870X2 Quadfire setup - that monster was over 300 watts at idle.

[JohnC]

71W idle, according to Kill-a-watt, not counting the monitor. Asus P8Z68-V PRO/GEN3 motherboard, 8GB of RAM, single 512GB SSD (I have external eSATA RAID storage, but it has its own power supply), i7-2600 (I don't overclock it), EVGA's GTX680 (stock clocks, I don't overclock it), WiFi network adapter. I never turn it off unless I'm upgrading drivers or BIOS or replacing hardware.

Edit: just tested it at maximum load, running latest FurMark (they had an update specifically for GTX680 cards) and encoding video in HandBrake - about 380W max. Not as bad as I've expected.

[Flying Fox]

In Ottawa the public library lends out Kill-A-Watt meters http://ottawa.bibliocommons.com/item/sh ... ill_a_watt
Check your library to see if they do the same thing

Wow, some library.

I bought my Kill-A-Watt on eBay in which the seller shipped to Canada. After shipping and everything I paid about $20 so I thought I had a good deal.