LinuxSociety.org


Practical Results of a PSU journey:


I run two computers with the following specifications:

(a) Workstation
CPU: Q9450 @ 2.66GHz 8x333
MEM: 4x2GB ddr2 @ 1000
VID: radeon 6670 1GB
OTHER: 3x hard drives + 1 dvd, audigy2, atx ip35-e motherboard
PSU: Antec earthwatts 650W (ea-650 early revision)
Pictures:
img img img img
(b) Server
CPU: E5400 @ 1.2GHz 6x200
MEM: 2x1GB ddr2 @ 800
VID: integrated intel
OTHER: 2x hard drives, matx g41m-es2l motherboard
PSU: Thermaltake TR2-430NP (sold at bestbuy and others)
Pictures:
img img img img

I purchased a SeaSonic X series SS-400FL for the Workstation.
Pictures:
img img img img img img img img img img
Here are the results:
Old vs. New
Item: OLD(EA-650)   NEW(SS-400FL)   diff %
IDLE(watts): 107W 97W -10.3%
LOAD(watts): 172W 157W -9.55%
3.3v rail: 3.12v 3.28v 5.12%
5v rail: 5.04v 5.12v 1.59%
12v rail: 12.02v 12.26 2.00%


I purchased a SeaSonic SS-300ET Bronze 300W for the Server.
Pictures:
img img img img img img
Here are the results:
Old vs. New
Item: OLD(TR2-430NP)   NEW(SS-300ET)   diff %
IDLE(watts): 69W 57W -21.05%
LOAD(watts): 83W 70W -18.57%
3.3v rail: 3.328v 3.360v 0.96%
5v rail: nm nm N/A
12v rail: 12.365v 12.556 1.54%


Conclusion:
Overall the results were excellent and i'm, so far, very happy with my psu upgrade. Seasonic quality seems superb, even on the budget SS-300ET.

Note: I took the highest noted measurement on all wattage readings with a kill-a-watt meter. The load was generated with a simultaneous running of an evolution simulator and a cpu/mem benchmark(both multi-threaded). GPU was not loaded. The idle numbers were taken prior to login after booting the systems. The voltage rail numbers were taken idle from the bios. The systems were clocked manually with SpeedStep disabled. The replaced units did not have any 80-plus certification(newer EA-650 revisions have a bronze certification now). The replacements (SS-400FL and SS-300ET) were 80-plus gold and 80-plus bronze respectively.

Overall summary of combined(both pc's) savings:
Item: Before       After       Diff Watts    Diff %
IDLE(watts): 176W 154W 22W -14.29.05%
LOAD(watts): 255W 227W 28W -12.33%


My main intent when I purchased these PSUs was to reduce power usage and get high efficiency units. I had predicted somewhere around 10% savings prior to the tests from the reviews i'd read elsewhere. I feel I've accomplished, or even surpassed, this goal.