Performance

Before getting to its overall performance, let's have a look at the voltage regulations and load ranges of this unit.  All data below was taken directly from the product manual included.

**click to enlarge**

All voltage readings were primarily captured with a quality True-RMS Digital Multimeter with data logging capabilities so that I could keep better track of any fluctuations easily and analyze the data better.

Also important to note is that all readings were captured directly from the 20/24pin PSU connector.  Attempting to capture readings from leads may not provide the most accurate readings. However, as I always like to stress heavily, be extremely careful when using this method and take all precautions.  The probes don't always stay in place well and typically will require support to ensure they don't not come loose.

Before getting to the results, let's have a quick look at the test system specs...

The HD3850 was powered via one of the four 6-pin PCI-E connectors featured on this power supply.

Now, the results here were a bit interesting, particularly in regards to a comparison between one of the last two Silverstone power supplies I've tested here.  First off, I like to log readings during the entire boot process.  Cheap power supplies will be unable to maintain steady voltages during this kind of testing.  While the ST70F is certainly not one of these, the numbers did jump around quite a bit and I wasn't really expecting this strictly based on my experience with some of their other units.  Looking at the numbers above, there was a .04 fluctuation on the 5V rail and a .07 fluctuation on the 12V.  The next logged performance tests were the readings at idle, captured after 30+ minutes of operation and logged for roughly a 15 minute span.  The results here were exceptional, with only a mere .001 fluctuation on all rails.  The final set of test that were logged were the readings under heavy load using both Prime and Sandra.  Again, I captured a 15 minute log during load using both of these tools.  I compared the results and numbers above are the absolute lowest and highest peaks from both tests.  Results here were not that bad, with the most significant being a .014 5V change and .012 difference on the 12V readings. 

While this particular unit did slightly struggle a bit when compared to their ST50EF recently reviewed here, note that unlike this unit targeted at a different kind of PC user, all were clearly on the higher spectrum throughout all testing, making these minor twitches insignificant to begin with.

Fan Noise

As far as fan noise goes, this unit will certainly impress.  This is a result of the new larger 135mm fan that operates at 18 dBA minimum.  I happen to have a system I manage that I personally built with one of their older Strider PSU's with the standard 120mm fan and can say that there is indeed a noticeable difference in noise levels.  Most importantly though, its excellent design does indeed aid in cooling and heat issues should be the least of your worries.  If you are looking for a high quality, ultra-quiet PSU, this one should certainly be on your short list.

Next:  Conclusions