Power Consumption: Stock & Overclocked
To compare the power consumption of the .09 micron Venice core based Athon 64 3800+ to the .13 micron Newcastle based CPU, we used a Seasonic Power Angel power meter to monitor total system wattage. Only the test system was plugged into the Power Angel, and the Power Angel was plugged into its own surge protector, which was plugged into an outlet that is on its own circuit. Our monitor was plugged into a separate surge protector, in a different electrical outlet altogether. To clearly demonstrate the effect the CPU was having on total system power consumption, we also took a few steps to ensure no other system components were drawing a significant amount of power during these tests. For reference, our test rig consisted of:
|Gigabyte K8NXP-SLI (NF4)
Corsair TWINX1024-3200XLPRO (2x512MB)
Western Digital 36GB Raptor
NVIDIA GeForce 6200 TC
Enermax 520 Watt PSU
3.5" Drive, Lite-On 16X DVD-ROM, Keyboard, Mouse
We configured Windows' Power Scheme to power down the hard drive after 1 minute at idle, and we disconnected the power cables from the floppy and DVD-ROM drives. We chose to use the NVIDIA GeForce 6200 w/ TurboCache video card because it doesn't require supplemental power, and is built using a minimal amount of components. Also note, that Cool 'n Quiet was disabled throughout our testing.
At both stock and overclocked speeds, the .09 micron Venice core based 3800+ used significantly less power than its .13 micron counterpart. With the .09 micron processor idling at its default speed, the test system used 10 fewer watts then when the .13 micron CPU was installed. And when running under a 100% load, the system used 30 fewer watts with the Venice based 3800+ installed.
While overclocked, the system obviously drew more current, but the Venice core based processor again used much less power. At idle, our test system used 12 fewer watts while idling with the .09 micron A64 3800+ installed and while overclocked it used 33 fewer watts.