 |
Swiftech H20-120 REV. 3 Liquid
Kit: When it comes to quality and performance, you can always
count on Swiftech. We look at their latest kit and give you
the 411. |
|
|
|
|
Test
System
| CPU: |
Intel
PIV 2.8E 775LGA HT Enabled
|
| Clock
Speed: |
2.8GHZ (14*200MHz) @ 1.3875v |
3.43GHz
(14*245) @ 1.5625v |
| Motherboard: |
Albatron PX915G Pro |
| Memory: |
1GB
Corsair XMS TwinX PC4000 Kit (2*512MB) |
| Memory
Speed |
3-4-4-7 |
| Hard
Drive: |
80GB
Western Digital 7200RPM SE 8MB Buffer
|
| Video
Card: |
Albatron
PCX5750 |
| Operating
System: |
Windows
XP Pro SP2 Direct X 9c |
| Drivers: |
Forceware
61.77 |
| Cooler: |
Swiftech
H20-120 Rev 3 |
Stock
Intel LGA775 Heatsink |
| Case: |
CoolerGuys Windtunnel IV |
| Power
Supply: |
RaidMax
400Watt Power Supply |
| Fan
For H20: |
Delta
WFB1212M 72CFM |
Vantec
Stealth SF12025L 53CFM |
| Test
Software: |
SiSoft
Sandra Burn-In 2004.10.9.133 |
SpeedFan
4.18 |
For our tests the system was run using the Burn-In test from
SiSoft for 30 minutes. We used the CPU and Multimedia
tests to make sure the system was running at full load with
both logical processors. After that we let the system
rest for another 30 minutes. All tests were run three
times with the closest of the tests being used.
SpeedFan was used to record the temperature to a csv file every
three seconds, which we then graphed onto a chart. Sound
Tests were taken from about 15" from the front of the case
with only the testing fan(s)/pumps enabled and the case closed.
So lets take a quick look at the sound levels from the system
with both types of cooling.
| |
Intel
Stock HSF |
Swiftech
H20-120 Rev. 3 Delta Fan |
Swiftech
H20-120 Rev. 3 Vantec Fan |
|
Sound
Level:
|
62dBA |
53dBA |
<50dBA |
We can see that the stock Intel cooler is loud, very loud.
However the Swiftech system with its default fan is very well
suited for each other, as the pump and fan are about the same
noise level. When we switched to the Vantec fan we noticed
that the pump was louder than the fan but both still managed
to stay under the 50dBA minimum of our test sound meter.
Stock
Clock Speed Cooling Results
First we will see how the two coolers compare at stock speeds,
on the power hungry Prescott CPU. All temperatures in the
test are in °C so any who are wondering why the temperatures seem
low, this is the reason. The ambient temperature was also
measured and placed in the graph, it never varied more than 1°C
in all our testing over multiple days. So lets take a look
at the Intel cooler first.

|
Intel
Cooler
|
Minimum
Temperature |
Average
Temperature |
Maximum
Temperature |
| CPU: |
28°C |
37.4°C |
46.5°C |
| Case: |
15°C |
15.8°C |
17°C |
| Ambient: |
|
12°C |
|

|
H20-120
Delta Fan
|
Minimum
Temperature |
Average
Temperature |
Maximum
Temperature |
| CPU: |
26.5°C |
35.4°C |
43.5°C |
| Case: |
15°C |
16.2°C |
17°C |
| Ambient: |
|
11.75°C |
|

|
H20-120
Vantec Fan
|
Minimum
Temperature |
Average
Temperature |
Maximum
Temperature |
| CPU: |
28°C |
36.2°C |
45°C |
| Case: |
16°C |
15.8°C |
18°C |
| Ambient: |
|
12.25°C |
|
So what can we see with these three graphs? We see that
the Intel cooler does well, never reaching over 50°C, though
in the ambient temperature that it is in it shouldn't as the
ambient temperature is 12°C, so any temperature over 50°C would
be well into the 60°C range in a 20+°C room. The water
cooler keeps the temperature below 45°C with either fan, which
is a good accomplishment considering the sound difference.
In a 20+°C room these two water cooler configurations would
barely be above the 50°C mark. Looking at the graphs the
stock cooler quickly drops its temperature when the load is
removed from the system. The delta fan based H20-120 takes
a little longer to cool the CPU down to its idle temperature,
but does so within a minute of the load disappearing.
With the Vantec stealth fan we see that the CPU takes about
two minutes to cool down to its idle temperature, which helps
to show that the extra added airflow helps cool the system down
quicker, at least at stock speeds and voltages. What happens
when we turn up the heat on these three setups?
Next
Page
|