Testing
Testing the H20-220 APEX Ultra involved a system from each of the major camps. In trying to keep up to date with current hardware I have sacrificed a standard for comparison. For example, I could have used the same Socket-939 Opteron motherboard combo so all water-cooling systems tested over the years would be compared to the same system. This would keep the wattage output of the processor consistant so long as Vcore, processor speed and all other components remain unchanged. The reason I have chosen to sacrifice this standard was very simple, I felt I can best serve you the reader by introducing new products at the height of their popularity. For watercooling I will test a processor under stock cooling, include another popular water-cooling system and the system reviewed. Hopefully this will aid those of you in the market for a water-cooling system (most likely the system featured) in answering how will the H20 kit tested improve (lower) CPU-temps compared to stock-air and another H20 kit. As a water-cooling comparison I chose the Corsair
Nautilus 500 external H20 system,
tested here back in July.
For processors I have chosen two fairly popular models from AMD and Intel. I have chosen a very overclockable Socket AM2 3800X2. This CPU dissipates 89W. For Intel I chose the Socket-T (775) E6400 which is also highly overclockable. While this processor only produces around 65W, its popularity warranted its inclusion.
Recently at Madshrimps we have eschewed IDLE temp readings from our result charts. This is because of the ambiguity in the term "IDLE." It's next to impossible to get a true "IDLE" or 0% workload reading since your CPU is
almost always performing some task no matter how indiscernible. To capture AMD processor "work load" activity I chose the program
AMD Power Monitor 1.0.2. (
Download). This utility is accurate enough to exemplify just how difficult it is to capture a 0% IDLE reading. Simply snapping a screenshot using the program
MWSnap the work-load lept from 0% to approximately 60%. Temps also increased concomitantly as high as 12C. Thumbnails below document these effects. Running a 3800X2 at 200FSB (2.0GHz) at 1.350V, one screenshot was taken for "IDLE" (left), the other at 100% LOAD (right). As the "IDLE" thumbnail shows I was unable to "snap" the screenshot without affecting CPU activity.
IDLE temps in the chart below were taken running the Apex Ultra on the fan's 5V setting which is near silent. I wanted to exemplify how close temps were when running IDLE with air-cooling on these processors. As stated earlier Swiftech includes fan voltage (speed) adapters for 5V, 7V or 12V operation. The latter does generate noise but even at the 12V setting this system is fairly silent. Running the fans on the 12V setting does little to effect IDLE temps given the fairly low heat signature of both processors tested.
I was impressed with the results seen below which were obtained running the fans at 5V. I tried the 7V adapters and found little difference in temps. However, running the fans full out at 12V produced an across the board drop in temps ranging from
3~10°C at 100% LOAD.
Sound Testing Noise levels were recorded using a Smart Sensor AR-824 SPL meter from a distance of 1-meter. Recording the noise level at different CPU speeds and/or varying LOADs would have little, to no effect on noise since the fans run consistently based on the chosen voltage adapter.
Fans at 5V = 34dB(A)
Fans at 7V = 36dB(A)
Fans at 12V = 42dB(A)
Epilogue
This review took much longer then I had expected since I have moved home recently to care of my mom. Personal issues aside manufacturers do expect reviews to be done in a timely fashion, however; Swiftech was very patient. Perhaps they are aware of the benefit to the readers when a product can remain in a system tested over an extended period of time. Living with a product allows me to provide you with a thorough review, flushing out (pardon the vernacular) any potential problems occurring over time and this is especially true with water-cooling. To be honest I have avoided 1/2-inch based H20 systems for almost two years due to the obtrusiveness of the "fire hoses" they require, the tenuous connection involved stretching a rubber hose over a steel barb then securing it with a plastic tensioner. And while many H20 Enthusiasts immediately replace these plastic fasteners with metal screw clamps, it seems ironic (or redundant) to go out and purchase "after-market" hardware on an "after-market" product? Fortunately hardware provided by Swiftech held up well.
PRO
Great water-cooling performance
Very silent with help of the 5V, 7V and 12V fan adapters
Versatile mounting and pump adjustability
CON
1/2" tubing is somewhat bulky (difficult to feed large hoses through case rear)
plastic fasteners should be replaced (this is, however; an industry wide issue)
Conclusion
The ergonomics of the desktop computer have evolved over the years and manufacturers have come to realize they must find a balance between form and function. Since its inception the H20 world was dominated by Overclockers satisfied with a utilitarian look since their primary concern had been performance. As this industry grows there are many more people water-cooling who must be concerned with what is known in the audiophile world as the WAF (Wife Acceptance Factor). The problem I have found with many 1/2" based systems, is they often come into conflict with the ergonomic ethos of the modern PC; miniaturization. In testing Swiftech's H20-220 Apex Ultra watercooling kit, found a balance between form and function, providing voltage adapters for silent running and the MCB-120 Radbox for versatile mounting. This kit was truly a pleasure to work with and highly recommended
Price: at the time of writing Swiftech's H20-220 Apex Ultra can be found for as low as $199 through Froogle.
Follow us on Twitter
Follow us on Facebook