Gone are the days when performance benchmarks settled all Intel vs. AMD processor debates.

The tone of the desktop computing world is palpably shifting towards power conservation, and even individual computer users are starting to ask themselves "just how much power is this going to use?"

Perhaps it's only when we consider the 24x7x365 electrical operating cost of a computer that these small power reductions muster any significance, but there it is. In a sense, what's happening now parallels the transition to from loud computers to the silent PC esthetic. Before hot-running computer processors went quiet, enthusiasts' attempts to reduce the PC noise were exceptions to the rule, rather than the norm.

An entire industry now thrives on selling quiet computer hardware, could the same hold true for energy efficient desktop processors that don't sacrifice performance? That certainly seems to be the direction electronics are headed, and frankly this is a much overdue change.

I think we're all still finding our feet on the energy efficient component, but it's certainly a new component PCSTATS will be considering as we mull over the benchmark results of the Sempron 3600+ processor. Now, let's start off with total system power draw tests.

Sempron Total Power Consumption Tests

The cost of energy is increasing world wide, both monetarily and environmentally. Any reduction in the amount of power a CPU requires is a welcome respite, but wading through the technical numbers to figure out the differences can be confusing because Intel and AMD report power values in slightly different ways.

On paper AMD appears to consume more power since its values are higher, however AMD lists its maximum power usage while Intel posts typical. Obviously the two are not comparable and to find out which actually consumes more power, we decided to test things out.

Representing Intel will be an Intel Pentium D 940, Pentium 4 840, and Pentium D 540 processor. Standing in for AMD are the socket AM2 Sempron 3600+, Athlon64 FX-62, Athlon64 X2 5000+, and socket 939 Athlon64 4000+ and Athlon64 FX-60 processors.

The test specs for each system stayed as constant as possible: Akasa PowerPlus AK-P550FF power supply, Western Digital 74GB Raptor hard drive, an 8x Gigabyte DVD-/+RW drive and a MSI NX6600GT-TD128E videocard.

Cool 'n' Quiet is disabled for these tests because with Cool 'n' Quiet enabled, total system power draw will be even less that what is reported. A picture of the total system power draw test configuration is shown at right. The Extech Power Analyzer is located between the mains 120AC supply and the PC's power supply.

PCSTATS uses an Extech model 380803 Power Analyzer Datalogger for measuring CPU power & total system power draw. The tool can be found here if you would like to replicate these microprocessor power measurements yourself.

To simulate single-core load, the Prime95 Small FFTs test was run, and two instances launched for dual core loading. Total system power consumption was read with a Extech Power Analyzer Datalogger (model 380803) for both CPU loaded and CPU idle states. These values are measuring an entire system while the processor is running under computational load or idle.

How does the socket AM2 Sempron 3600+ impact total system power draw? Let's have a look....