Related Articles

We were quite surprised and disappointed to find that the Thermaltake Silent PurePower did not fit in our standard ATX chassis. The illustrations in the manual depict a square hole in the rear of the chassis to allow for the cooling block, whereas our chassis had a round hole to accommodate the standard fans on the rear of power supplies. As a result, the power supply was mounted too far left when facing the rear of the chassis, and while it was fairly sturdy (it wouldn't fall out in a stationary system), we definitely wouldn't want a power supply just hanging there with no screws holding it in place. This is definitely a limiting factor in the amount of chassis that can be used with this power supply, and as such, should be factored into your purchasing decision. Sadly, there is no mention of this important fact in the small manual supplied with the Thermaltake Silent PurePower. Doing a quick look around the office at a dozen cases showed that ten of them could house the PurePower without a problem - most had a larger opening than our particular test case. Most people should not have this problem but it may be worth checking out the back of the case just to make sure.

That said, it wouldn't have been too difficult for Thermaltake to design the heat sink block as a cylinder, rather than a cube, that would fit through the round hole cut out for the 80 mm power supply fan. This would allow the Silent PurePower to be mounted in virtually every ATX chassis.

Other than that one problem, the installation of the Thermaltake Silent PurePower would have otherwise been the same as every other power supply. As mentioned before, the cables are a good length for mid to full-sized ATX chassis, and did not get in the way most of the time.

The performance of the Silent PurePower is a mixed case, unfortunately. We tested it on our in-house Athlon 64 bench system, and while the power output levels remained at acceptable limits, the temperatures of both the CPU and system were much higher when idling and under load than when using a regular power supply. Our test system was as follows:

• AMD Athlon 64 3400+ (Socket 754)
• Soltek SL-K8AN2E-GR Motherboard (nForce 3 250Gb)
• 2x 256 MB OCZ PC3500 EL (Dual Channel)
• Seagate Barracuda SATA V 120 GB
• eVGA GeForce 6800 Ultra 256 MB/BFG GeForce 6800 GT 256 MB*
• Gigabyte 3D Cooler-Ultra GT (GH-PCU31-VH)
• Standard mid-size ATX chassis
• Thermaltake TWV-480AD Power Supply (double fan)**

** This was our power supply for determining system and CPU temperatures in a normal situation (non-fanless).

We ran Motherboard Monitor 5 in the background at a monitoring interval of 3 seconds, while looping through the game and CPU tests of 3DMark 2005 for about 35 minutes to stress the power supply with the components we had in the test system. The results of idle were recorded 30 minutes after the 3DMark looping was aborted. The results are as follows:

Voltage Rail Performance

Overall, these values do not stand out, especially when compared to some other power supplies on the market, but they are by no means bad. Keep in mind that even the seemingly-low 11.80V under load is still very much within the commonly-accepted +/- 5% fluctuation that is acceptable for the ATX specification.

An item worth mentioning is that the GeForce 6800 Ultra is quite power-hungry, as can be deduced by the fact that it requires two Molex connectors. The load that is put on this power supply would differ when used in a regular system, as most people are probably not lucky enough to have the latest and great video card. On the other hand, many of us have multiple hard drives and optical drives, which draw a fair share of power as well. There is no reliable way to determine the kind of loads that may be put on a power supply when it is being used in a regular environment, so therefore we try our best to stress the unit with what we have available.