|Zalmans Noise Prevention System|
The 5100 is a strange creature...it features Zalmans Noise Prevention System, which basically boils down to running the fan at voltages less than 12V. Included in the box are two fan header cables, which incorporate two different levels of resistors, which take the current down to around 7V and 5V. The FHS will only mount on boards which have 4 holes around the socket - this means all Pentium 4 boards, and some socket A (including the Abit KT range). This makes installation more...interesting, shall we say :)
The heatsink. Rather chunky, eh? Looks like a spaceship or something...the thing weighs a ton. I was a little apprehensive about mounting it on my board...guess that's what the mounting holes are for.
The base is machined nice and flat, to ensure a good surface contact. The flower design is certainly unique, and each fin is contoured for a better airflow. The fins tend to rip at your skin a bit though, you should see the state of my fingers now!
Back to the serious stuff...Unless your case already has mounting supports installed (highly unlikely) you're going to need to remove your motherboard and install some screws underneath. Zalman supply washers to stop you scratching your board, which is rather handy. the screws stick up from underneath the board when re-installed...
So, on to the good stuffs :) Unfortunately, I don't have an entire army of Socket A coolers here to compare it with, so let me explain my testing methology a little. My personal opinion, when I read a HSF review, is that I don't care how well a cooler performs in someone elses system, I want to know how well it will perform in MY system. Therefore, I always like it to be compared to something standard, so that I can see what sort of a difference it will make. Now, I like to use a FOP32 to benchmark against. This is because nearly every socket A user, at some point or another, has used a FOP 32 or 38, and knows how well it performs in their system, and knows how well their current cooler compares to it. It's something of a common denominator, so by comparing the FHS to a FOP, I hope you can estimate the kind of difference it might make in your system.
Firstly, a note on the NPS (Noise Protection System). This is amazingly effective. I considered putting my last HSF on a Baybus, and running it at 12v/7v , but never really had the guts, due to immense fear of burning my CPU up! It's good to see a cooler manufacturer run with this idea. There are 3 modes - the fan can be connected straight to the mother board (running at full power and full noise) or can be connected through 1 of 2 cables, running at either low noise or silent mode. They do exactly what they say on the tin - low noise is extremely quiet, and silent is, well, silent. You'll be amazed how unnerving it is to start up your machine and not hear the processor HSF!! I love a quiet system, and this HSF is absoloutely perfect for this. However, it does take a toll on performance, so see below....
So, we can see that the full power FHS performs rather well. On my Duron 800@1000, it cools the idle temperatureto 28, the same as the FOP. All the idle temperatures are within a few degrees of each other, but it's where the CPU is under load we see the difference. For the load tests, I ran Seti at full whack, and left an MP3 playlist running just for good measure, which should completely utilise the CPU (anything the MP3s don't use, Seti will). The full power FHS performed very well, keeping the CPU at a cool 41 C. We can see that as the cooler gets quieter, it's performance worsens - silent and low-noise mods are both out-performed by the FOP, which is rather poor. However, I cannot emphasise enough how quiet these are!