Testing is a check of the different rails with the PC at different levels of stress. At idle, load, and overclocked load.
The test setup is as follows:
CPU | Athlon 64 3000+ |
Motherboard | Abit AN8 Ultra (Bios V1.6) |
GFX Card | X550 256MB |
Sound | Realtek ALC850 |
CPU Cooler | |
Main Memory | 2 x 512 MB Corsair XMS PC4000 |
Non-removeable Disks | 1 x 80GB Western Digital SE, 1 x 120GB Maxtor SATA |
Removeable Disks | 7-in-1 USB Flash drive |
Optical Drives | NEC x16 dual-layer DVD-RW |
Case | PC-Case Nitro AX |
PSU | UNDER TEST: XG Magnum 500W |
The processor and RAM will be overclocked to 240FSB, staying with a x9 multiplier on the CPU for 2.16GHz @ 1.5V. The RAM will be at 240MHz (480MHz DDR), 3.0V with 2-3-3-8 1T timing.
Load is running Prime95 Torture test, Sisoft Sandra CPU Burn-in, and 3d-mark 2003 looping for at least 30 minutes, then taking all reading for that setting.
The voltages will be tested using a multimeter, while the wattage output will be read from the display on the PSU.
Output | Idle | Load | Overclocked Load |
+5V | 5.17 | 5.16 | 5.15 |
+12V1 | 12.11 | 12.09 | 12.08 |
+12V2 | 12.20 | 12.15 | 12.05 |
+3.3V | 3.39 | 3.35 | 3.35 |
Total Wattage | 84 | 97 | 105 |
This is a rock solid supply. Also the temperatures only went up after the overclocked load test. This is a good indication about how efficient the powersupply is. I am not certain about the 105W reading, but it could be accurate with that graphics card. Even with Dual 7800GTX and an FX-57, I have my doubts that more than 400W would be sustained for a long period of time.
The temperature only reached 38C, which is the same as the Sys temperature from the motherboard diode. The external heatsink gets quite warm, but not hot. It works well to keep the PSU cool. The fan did not engage at all during the testing, so for a medium system, it is totally silent.
Next Page: Conclusion