Installation

Installation was easy, as I wasn’t installing into a case, but using it with our test bench PC. However, the additional length of the unit may make it a necessity to remove your motherboard in order to slip it into your case instead of carefully bending and forcing it into place. The unit is remarkably and reassuringly heavy, but may cause issues in cases where the only support are the four rear screws.

Testing

Methodology

As we’ve yet to come up with an expandable, real-world, repeatable test for analysing power supplies it should be noted that these results can only be compared to the models listed below and not with any of our other reviews.

For this test, we’ll be using the below configuration.

Processor	Intel E6750 Core 2 Duo @ stock
Motherboard	XFX 780i Tri-SLI
RAM	4 x 512MB OCZ DDR2
Drives	Western Digital Protégé IDE HDD SATA DVD drive
Graphics card	XFX 8800 Ultra XFX 8800 Ultra

While we have the ability to test tri-SLI, none of the power supplies had the required connectors in order to feed the three cards (four PCI-e, not six); hence this is the current maximum GPU load these units will ever see.

To test, we run full screen (1600x1200) RTHDRIBL which is a primarily graphics stress program, although it uses 50% of the CPU too. This is run for half an hour and the voltages are taken, along with the PF reading, and the wattage pulled from the mains. The idle testing is done with nothing running and the PF value and wattage are read again. The system is tested with a run-through of 3Dmark06 to make sure that the units are stable, and to test the noise output.

Results

Click to enlarge

From these results, the power supply was at a third load (400W peak, 1200W unit) and managed a PF value of 0.91 at load and 0.88 at idle. The PF value relates to the efficiency of how the unit copes with switching the AC input to DC output; read more here.

Power supply	PF idle	PF load	Watt idle	Watt load	Watt standby	Cost (24/7 idle @ 7p kWh)
FSP Everest 1010W	0.92	0.94	273W	400W	2W	~£167
Jeantech Absolute 1000W	0.94	0.96	277W	406W	3W	~£170
Tuniq Ensemble 1200W	0.88	0.91	274W	403W	2W	~£168

Between the three units, the Ensemble fits in between the Everest and the Absolute.

As you can see from the above results, none of the power supplies tested broke the 5% tolerance levels that the ATX specs set out. While the Everest is most faithful to the ideal voltages, it varies the most, but this is in comparison with units which are much larger and the differences are in the order of hundredths of a volt.

The Ensemble does a sterling job at keeping the rails stable, with variation across the board of only 0.01v between idle and load regardless of voltage, although its voltages are consistently above the required level with its 5v being up to 0.16v away from the ideal.

It should be noted that this differences are tiny representing single digit percent changes in voltage, and won’t affect PC operation as a result; for the most part stability is more important than getting the exact voltage.

The noise output of the unit was on par with the other units, although it was mildly quieter than the Everest thanks to the larger fan. The 80mm fan makes it a tad bit more audible than the Jeantech, but these differences are unlikely to be noted once installed into a case. Another plus point of the unit is that it doesn’t make the whirring electrical sound when put under load which is noticeable with the Jeantech unit.