 |
Before I get into any form of actual performance testing, one factor which is important to look at is the overall power consumption of a system. By using high draw components such as a massively overclocked Prescott, 2GB of Performance memory, and a fast videocard, it's rather easy to make the system pull a few hundred watts from the wall. The question then becomes which supply can pull the fewest watts, yet still provide the same power?
Stock Speeds
Idle Draw
| Enermax Liberty 500W |
148.5, Power Factor .990 |
| Cooler Master eXtreme Power Duo 650W |
153.4, Power Factor .664 |
| Mushkin Enhanced XP650 650W |
159.2, Power Factor .630 |
Full CPU/Memory Load
| Enermax Liberty 500W |
210.3, Power Factor .993 |
| Cooler Master eXtreme Power Duo 650W |
214.9, Power Factor .686 |
| Mushkin Enhanced XP650 650W |
220.5, Power Factor .657 |
Full CPU/Memory Load + 3D
| Enermax Liberty 500W |
236.4, Power Factor .996 |
| Cooler Master eXtreme Power Duo 650W |
244.0, Power Factor .675 |
| Mushkin Enhanced XP650 650W |
247.5, Power Factor .666 |
Overclocked Speeds (CPU @ 3.83, 1.65V)
Idle Draw
| Enermax Liberty 500W |
191.0, Power Factor .992 |
| Cooler Master eXtreme Power Duo 650W |
195.1, Power Factor .674 |
| Mushkin Enhanced XP650 650W |
199.1, Power Factor .635 |
Full CPU/Memory Load
| Enermax Liberty 500W |
288.9, Power Factor .995 |
| Cooler Master eXtreme Power Duo 650W |
300.0, Power Factor .700 |
| Mushkin Enhanced XP650 650W |
297.8, Power Factor .632 |
Full CPU/Memory Load + 3D
| Enermax Liberty 500W |
330.0, Power Factor .996 |
| Cooler Master eXtreme Power Duo 650W |
332.0, Power Factor .707 |
| Mushkin Enhanced XP650 650W |
344.0, Power Factor .710 |
As it's shown, the inclusion of Active PFC on Enermax's Liberty unit (also as shown on my APFC or Not? Review in past) really helps to even out the incoming power, and as such, results in marginally lower consumption. The one thing which sits strangely here is that of the three units, the Enermax has the coolest operation, as well as the lowest cost. As things panned out, it seemed that the Mushkin unit gave the end user access to the most power (a combined 44A on the 12V rail when neccicary), but on the same token, consumed marginally more power than the rest of the pack under loads. With power consumption put down for the count, it's good and time to see how voltages fluctuated under load.
|
|
