Power consumption

[muon]

How does cpu power consumption scale with fsb when compared with multiplier changes.

Does a C2D with 500x6 use any more or less when compared to 333x9. Do they require different voltages?

Ignore any chipset power consumption.

 

[neodude]

I wouldn't have thought it would make any difference.

 

[jeffa123]

I've seen overclocking at lower voltages especially with the Q6600. Personally I believe the power consumption is dependant the load that is place upon it. But also remember If you have a reasonably good Graphics card in your PC you have also got to factor that in. A Hard drive will consume 60w just being on. A Graphics card will cost you 150-350W alone. The MB will start at 35W and then there are the fans or cooling system you choose. Then there is the PSU. The average PSU used on a PC now has to be around 700w. And slowly they are all going BTX instead of the old ATX. The BTX design is a PSU which controls and monitors its own power draw so would be a better choice than an ATX because it is much more efficient and draws only the power it requires. So if anyone says about how much power you will save, think about how power consuming your current setup is before making a decision.

 

[pcoltrane]

I've seen overclocking at lower voltages especially with the Q6600. Personally I believe the power consumption is dependant the load that is place upon it. But also remember If you have a reasonably good Graphics card in your PC you have also got to factor that in. A Hard drive will consume 60w just being on. A Graphics card will cost you 150-350W alone. The MB will start at 35W and then there are the fans or cooling system you choose. Then there is the PSU. The average PSU used on a PC now has to be around 700w. And slowly they are all going BTX instead of the old ATX. The BTX design is a PSU which controls and monitors its own power draw so would be a better choice than an ATX because it is much more efficient and draws only the power it requires. So if anyone says about how much power you will save, think about how power consuming your current setup is before making a decision.

LOL - where do you get these numbers from?

60W for a hard drive? 350W for a graphics card?

A PSU may well be rated to deliver 700W but the actual power draw will only be that of the sum of the components in the system plus 15-30% dependent on the PSU efficiency. I think you are a bit lost with the whole ATX/BTX thing too!

To give you a more realistic idea - All my HTPC pcs draw between 50W idle and 90W full load, measured at the wall (therefore including the PSU inefficiency)! Even my main rig only peaks at about 220W

Going back to the OP - to give you a rough idea, the power used by a processor will increase linearly with clock frequency, but will square with any voltage increase. It's actually a little more complex than that but it will give you a good idea.

Eg ... 1ghz -> 1.3ghz = 30% increase in power
1v -> 1.3v = 69% increase in power

As for FSB vs multiplier ... in theory is should be about the same amount of power draw for the chip, but with a higher FSB, as the motherboard is running at a higher frequency there would be a slightly higher power draw associated with that.

 

[weescott]

60w for a HDD? lol. typical is around 18w.

A higher FSB clocked chip might consume more power through the chipset, but it's not going to be much.

 

[siege2]

u could get a Plug-In Mains Power and Energy Monitor

Specifications:FunctionRangeAccuracyTypicalMaxRMS Voltage180-250V rms0.2%1%RMS Current0-15.00A rms0.3%1%Active Power0-3750W0.5%2%Line Frequency47.0-63.0 Hz0.1Hz2%Power Factor0.00-1.000.010.03Power Consumption20W max

a well known Electronics UK company sells them all over the UK, Metro Centre being one, sites all over the UK..

 

[willhub]

Graphics cards consume around 20-300W depending on the cards and how many cards? Dont the 8800GTX use around 250W?

 

[muon]

I know about pc power consumption and how it scales with clock speed and voltage. but thanks anyway.

However, my question is pretty specific. I did actually get a power meter yesterday. When i have some time i'll be sure to post anything i discover. I wont be able be able to eliminate northbridge consumpion differences between the two setting, but by using different memory multipliers, whould be able to eliminate any bias due to that.

The issue with the northbridge is that, i can either assume that northbridge power conumption is constant and measure the difference between the cpu multipliers, but that would be incorrect. On the other hand, i can assume that cpu multipliers dont affect cpu power consumption and thereby measure any northbridge consumption differences by changing the fsb. Which basically mean i wont actually be able to give a definitive answer.