PDA

View Full Version : PC Electricity Consumption?



worldpease
02-11-2005, 07:49 AM
how many watts does a PC use?
for example by hour,...
how many watts does it use by hour, when I leave it on by night?
(considering that is just the CPU unit, not the monitos).

uNz[i]
02-11-2005, 09:29 AM
Here (http://greenstuff.decoder.com.au/tips.htm#16) is a list of PC parts with how many watts each part uses.

Edit: It's a bit of an old list though... but it'll still give you a general idea.

lynx
02-11-2005, 11:43 AM
Nice list.

Remember that most of those are PEAK power consumption figures, generally the use will be a fraction of what is quoted there.

worldpease
02-11-2005, 07:28 PM
Here at home,
they blame the high bill rates on me, they say Im the one with the computer, and that I leave it all night... but I don't think so, the CPU can't take more energy than the freaking lights that no one turns off, the washing machine that they forget to turn off, or the 25" to 27" TVs that are on at least 8hrs a day.
I guss I'll just have to pay the bill and see if they leave me alone for once.

tesco
02-11-2005, 09:37 PM
lets jsut say my computer was to work at full power consumption for 24hours a day, 7days a week for 1 month.
450watts(psu)+35watts(monitor)=485watts
485/1000=.485kw
at 5cents/kw electricity that would be:
.485*5=2.4cents/hour
2.4cents*24hours=57.6cents per day
57.6cents*30days=1728cents (17.28dollars)

so to leave teh computer on 24/7 it would cost ~17dollars per month! :D

now i most likely did something wrong here but oh well. :P

Virtualbody1234
02-11-2005, 11:41 PM
Ross.

The 450 watts rating of a power supply is the total output. Not it's consumption. :no:

tesco
02-12-2005, 12:15 AM
Ross.

The 450 watts rating of a power supply is the total output. Not it's consumption. :no:
oh.
so what does it consume then? :ohmy:

edit: did i get the other part though? like would the 450watts be watts used in each hour or something else?

lynx
02-12-2005, 01:45 AM
The PSU is probably about 80% efficient, so the total consumption should be about 600W not 485W. Other than that your calculations are correct, you just need to increase the figure by about 25%, making the total (based on 5c/KWh) about $21.50 per month.

Of course, you wouldn't actually run so close to the PSU's maximum capabilities and in any case you could never run every component at full power 24/7. So you could probably make a guess at the actual consumption being about a quarter of that, or about $5.50 per month.

Virtualbody1234
02-12-2005, 03:24 AM
Consumption is what is drawn from the wall outlet. Not the output of a power supply.

I did a small test. I hooked up an Amprobe similar to this:

http://www.torontosurplus.com/tes/DATA2519.JPG

With my PC running at idle it shows a draw of 1/3 of an amp.
Under load (Folding@home) it shows a draw of 1/2 of an amp.

And I'm being generous. It shows actually less than these figures.

120 volt line @ 1/2 an amp = 60 Watts.

So in fact my PC is running under full load and the consumption is less than 60 watts (not including monitor).


So... 60w x 24h x 30d = 43200wh

43200wh / 1000 = 43.2 kwh

@ 5¢ per kwh that's:

$2.16 per month or $25.92 per year.

worldpease
02-12-2005, 03:39 AM
Damn, you guys are good.
I don't think it makes diference, they just won't "see".

tesco
02-12-2005, 03:41 AM
With my PC running at idle it shows a draw of 1/3 of an amp.
Under load (Folding@home) it shows a draw of 1/2 of an amp.
Why is your's so efficient? Lynx said they're 80% efficient but by your results they're way more efficient (over 100%, im too lazy to caclulate it).
You sure those are right?
how could the full load be less than the idle?

tesco
02-12-2005, 03:44 AM
oops

Virtualbody1234
02-12-2005, 03:51 AM
Why is your's so efficient? Lynx said they're 80% efficient but by your results they're way more efficient (over 100%, im too lazy to caclulate it).
You sure those are right?
how could the full load be less than the idle?

Why do you keep refering to the power supply output or a percentage of that?

We're talking about consumption here (input).

tesco
02-12-2005, 03:56 AM
Why do you keep refering to the power supply output or a percentage of that?

We're talking about consumption here (input).

The PSU is probably about 80% efficient, so the total consumption should be about 600W not 485W.

that or i'm really not understanding something and you should explain. :frusty:

lynx
02-13-2005, 01:53 AM
A 450W PSU is theoretically capable of supplying 450W. If it did, it would consume about 550W (assuming 80% efficiency). But as I said, you don't run the PSU at anything like full power unless you are asking for trouble.

I made a guess at the actual consumption being about 25% of capacity. I knew I was probably on the high side, but I wanted to show the worst possible scenario.

High graphics activity(playing games), high levels of disk activity etc, can probably add a lot to VB's "folding" consumption, maybe as much as double with some graphics cards (but that's a guess) which would bring it in to line with my figures.

In any case, both our figures are an approximation, and in engineering terms a ratio of about 3:1 is generally regarded as being in the same "ball park".

Shiranai_Baka
02-13-2005, 02:39 AM
What is your setup VB?

Virtualbody1234
02-13-2005, 03:05 AM
What is your setup VB?
http://www.filesharingtalk.com/vb3/showpost.php?p=990014&postcount=12

Also connected to the power supply:

1x 12" Cold Cathode light
4x LEDs connected in series to the 12v line (3volt each).

Snee
02-14-2005, 02:07 PM
']Here (http://greenstuff.decoder.com.au/tips.htm#16) is a list of PC parts with how many watts each part uses.

Edit: It's a bit of an old list though... but it'll still give you a general idea.
When it comes to top off the line cards I think you can double the wattage for agp cards given in that list, and PCI-e draws just as much.

I think on-board peripherals (lan, sound etc) will draw slightly less than their pci counterparts.

Big HDs today pull more than what is mentioned there I think, while some AMD processors (mobiles) will consume less.