The test setup
Now down to business. Performance is a major factor when buying a water-cooler. A second factor is the possibility for noise-reduction. I strongly believe that a water-cooling system should be capable of mastering both in order to justify its price tag and also allow for some kind of ability to control what exactly you want to do. Otherwise I’d find paying more than $100 for a cooler difficult to justify for the average consumer. As I am particularly fond of silent PCs I paid particular attention to its noise-level, but also tested its limits in terms of thermal load.
The test machine:
- AMD XP2400+ CPU
- Epox 8K5A2+ motherboard
- 1x 512MB Corsair PC3200 LL DDR
- 3x Seagate Barracuda IV 80GB 7200rpm
- ASUS V8460 Deluxe GeForce4 Ti4600 AGP
- CWT 480W "Noiseblocker" PSU
- SoundBlaster Live Player
- 3Com 3C905C-TX-M 10/100 NIC
- 4x/2x/8x DVD-RW Drive
- 16/40x DVD Drive
- Windows 2000 Professional
- All in a ChiefTec big-tower with 2x 80mm noise-blocker fans at 1050rpm.
What I went looking for were the CPU temperatures when under heavy load while running the radiator’s 120mm Sunon fan at 12V and at 7V. I also over-clocked the system to the max to see if the lower temperatures made much difference in terms of overclockability. As a point of reference I will use the air-cooling I have which consists of a silent CPU-cooler and 5x 80mm noise-blocker fans running at 950rpm which produce less than 17Db(A) each.
Thanks to my brand new air-conditioning I kept my room at a steady 24C, which is a lot better than the 30C to 32C I would otherwise be facing. Hence my familiarity with home made super-cooling setups.
The temperatures are right where I would expect them to be. The WaterChill™ gave me significantly lower temperatures than my air-cooling and I would wager that you would be very hard pushed to beat those temperatures with anything less than a 12" high-speed fan blowing air into the side of the case.
I do not recommend passive cooling for the WaterChill™ though, as this yields temperatures higher than a decent low-noise Heat-Sink/Fan setup (63C at 2000MHz). Due to the availability of fairly quiet Heat-Sink/Fan setups for under $40 there isn’t really any benefit from doing so.
- 15x100MHz = 1500MHz at 1.60V
- 15x133MHz = 2000MHz at 1.65V (default)
- 15x150MHz = 2250MHz at 1.75V
All temperatures are at full load.
At 2250MHz the air-cooling is over its limit. A small rise in room-temperature would send me down in a blaze of blue glory. The WaterChill™ has no problems at this speed. All the temperatures with the WaterChill™ are significantly better than air-cooling.
Please note that the air-cooling setup used here is likely to be better than one found in pre-built systems and to the fact that there are numerous other devices contributing heat to the system as a whole.
I have to say that when running the 120mm fan at 7V the setup is moderately quiet. My air-setup is still quieter though, even with the case open, which gives me more AGP over-clocking possibilities. In defense of the WaterChill™ I have to point out that my air-cooling setup involved untold hours of work finding the right equipment and the right settings. I paid over well $100 for everything and at least 10 hours of work installing as well as running to my local retailer a dozen times and have at least 10 80mm fans and 3 Socket-A coolers sitting in my drawers because they are too loud or not powerful enough. The WaterChill™ kit was no effort at all and after two weeks is still as quiet as on the first day. An additional bonus is that Asetek have GPU and Chipset coolers which can be used in combination with the CPU cooler, so I know my Ti4600 is going places in the near future.
At 12V the WaterChill™ is as loud as a regular, well-endowed Heatsink/fan combo, but will cool an awful lot more. Due to the steady speed of the 120mm Sunon fan the pitch and noise doesn’t continually change as is the case with many CPU fans which I find extremely irritating and distracting when working, but the fan does develop a very slight whine at 12V so you may have to position it properly to avoid your casing amplifying that noise. By positioning the Radiator a little further into the case I was able to reduce the noise-level and eliminate that whine.
Please bear in mind that my case is fitted with noise-dampening material and that if your case lacks this then you may encounter slightly different results with noise-levels when positioning the radiator.