For Snowcrash and anyone else that might be interested.

The ATI HD4890 is a large 256bit card that under full load draws quite a lot of power, to help get over the high power consumption, particularly when the card is at idle, it automatically underclocks its GPU but it doesn't underclock the memory, that is a little unfortunate as the RAM requires plenty of power to run at its default speed of  975Mhz.

So how much power can we save using the ATI Tray tool and underclocking both the GPU and RAM to a speed that will be fast enough for every day mundane PC tasks including full HD video playback in Windows 7.

The answer is 46 Watts.

NOTES:
The power readings are for the entire PC (tower case).
The readings are at idle, no heavy graphics application was running.
The canter reading was the card automatically underclockling its self.   

Thanks for that. Good to know.

A link back to power savings on the processor.
http://www.datahopa.co.uk/forum/pc-hardware/underclocking-an-intel-q9400/msg8908/#msg8908

So, 27-37W saving on the proceassor.
and 46W underclocking the graphics.

85W average, worth saving.

Yeah and 85watt over a year is equal to something like £68

Just underclocking the graphics card is going to save about £40 a year, minus the time I spend gaming.

I'm curious as to how much power the entire setup consumes, including the monitor(s), printer, and any other devices that are used. If they're all plugged into a single power strip, that should be an easy measurement. Or is that how you were measuring it?

Dave, your wish is my command, here is my bat cave and power consumption with PC Underclocked and @ idle.

     37 Inch Screen.

     PC Tower

     External HDD

     5.1 speakers
 
     Router

    Cable modem.


Hard to see from the picture but its 208watt.


Beat that with a 37-inch screen  


[attachment deleted by admin]

Sadly, I don't have the means to measure AC current for my system, as I don't have that type of meter. I will, however, list my system components, so you have an idea of the (max) power requirements. I just have to go inside to do it, so it will take me a few minutes. Please stand by for more info.

Ok, here we go:
   
System specs:                      Power Rating

Computer (PSU)                       550 watts         

Monitors:
1.) Visio VW32LHDTV20A HDTV   180 watts
2.) Dell E197FP                         40.6 watts
3.) Dell ST2220M                       28 watts

External Storage:
1.) Iomega Z100USB zip drive      ~5 watts*
2.) HP Simple Save 1TB HDD        ~10 watts*

External Audio (LabTec LCS-1060)~20 watts*

Total (max) Power Consumption      833.6 watts

* Approximate, based on ratings of the attached power supply

Your power supply reading will throw your equation way off, I have a 600Watt PSU but the PC power consumption is only between 100 and 200watts, yours is probably in that ball park too. 

The bit that will take some beating is my screen, I have the backlight turned to minimum, which is perfectly bright enough and that takes the power down to 75Watt from 130watt, there aren't that many screens that allow the turning down of the backlight.

A new LED backlit screen might beat it though, inch for inch. 

I'm not quite sure that the power consumption will be that low for the PSU, in all honesty (though I'm quite sure it's nowhere near the max). It powers 5 physical HDD's, 2 DVD-ROM drives, the graphics card, 4 high capacity cooling fans (2x2.5" and 2x3.5") and 3 standard case fans, as well as the Mobo. I really wish I had the means to measure total output from the computer, at the very least. Maybe I'll break down and go buy an inductive wattmeter.

HDD's take about 8 watts each, fans take very little, say 1watt each, not sure about DVD drives to be honest, I guess they take more when they spin. 

And your graphics card is a lower end card, so it shoudnt take that much power.

I want to know what its taking now 

I know this link is no go for you Dave but this is almost exactly the device I use to test power consumption.

http://www.amazon.co.uk/Plug-In-Power-and-Energy-Monitor/dp/B000Q7PJGW/ref=sr_1_1?ie=UTF8&qid=1338917115&sr=8-1

That's a nice tool there, Data. What I would be looking for is something more like this. Granted, it doesn't measure wattage directly, but it DOES measure both voltage and amperage, and if you multiply the two together, you get wattage, according to Ohm's Law (P=I*V - where P is power {watts}, I is current {amps} and V is obviously voltage). And at $12USD, it's affordable enough, but I've the transit of Venus going on (through VERY cloudy skies, darn it!), and have my Grandmother to care for besides. I'll see about zipping off to Harbor Freight tomorrow to pick one up.

I have a multi meter similar to that one, please don't feel that you have to get it just to satisfy my curiosity Dave.

I do use my one quite often, the most common use I find is testing batteries.

I can also imagine that it might be tricky to test the power consumption with one of them, if you do get it please be careful my friend  

Data, I love you dearly, and I appreciate your concern. However, I don't think there's much danger involved, given that I've a goodly number of years worth of experience in dealing with taking electrical measurements in far more dangerous situations than simply placing an inductive probe around an insulated cord. My computer's battery backup already provides a precise measurement of the incoming voltage from the wall, so the only danger involved here is tripping over the jumbled mass of clutter that surrounds the floor in front of my filing cabinet while on my way to gaining access to the power cord to my computer. That said though, I'll be sure to clean my room before taking any measurements that may cause me to knock over my collection of Rubic's Cubes.

A good use of terminal blocks should give a lovely tight connection before power on.  *** insert Bzzzz sound ***

Thanks for the vote of confidence, Snowy!

Stopped in at Harbor Freight, and found a power consumption meter similar to what you've got, Data. I just plugged my laptop into it, and it seems to pull 24 watts at 120 volts and 0.33 amps. Sadly, the numbers don't seem to be matching up, which means that one of the measurements is incorrect.

If V = 120 volts, and I = 0.33 amps, then P needs to be 39.6 watts, not 24.

Whatever. It was cheap, and I can measure the voltage independently, for verification purposes. I'll plug it into the main box later tonight, to see what I get. Then I'll post some pics.