Even unoverclocked Athlon CPUs can consume up to 54 Watts (650 MHz Athlon) - and Athlon CPUs overclocked to 700MHz or more, or using core voltages higher than 1.6V, will gobble up even more power. For this reason, CPU cooling is an important factor to consider when building an Athlon system - no matter if you have overclocking plans or not.

Reason enough to take a closer look at high performance CPU coolers for Athlon CPUs. We've tested 10 different coolers - most of them being dual fan units.

The testing methodology for this test has been described in the heatsink testing methodology article.

Let's have a look at the test results first:

Test under real-world conditons

Heatsink 
    Temperature
Global WIN VOS32-plus
(1)
41
28
Global WIN VOS32
44
32
Alpha P7125
YS Tech fans
45
32
Alpha P7125CM60
Sanyo Denki fans
46
32
Montac ArcticCircle
48
32
TennMax VIVA STF
52
32
RDJD K701
52
32

  CPU temperature   Motherboard temp.
All values in °C.

Tested are only the heatsinks that would fit the test setup (motherboard: Asus K7M rev. 1.04)

(1) The VOS32-plus is a heatsink/case cooler combination with fan duct.

Test with "simulator"

Heatsink 
    Temperature
Global WIN VOS32-plus
(1)
n/a
Alpha P7125
YS Tech fans
31.8
Global WIN VOS32
32.1
Alpha P7125CM60
Sanyo Denki fans
33.7
Global WIN FKK32
34.4
Montac ArcticCircle
(2)
36.6
Global WIN FKK50
37.6
Global WIN VEK32
40.1
TennMax VIVA STF
40.7
RDJD K701
40.9

  CPU temperature   Motherboard temp.
All values in °C.

 
(1)The VOS32, being a heatsink/case cooler combination, cannot be tested with the simulator.

(2)The Montac ArcticCircle, unlike other heatsinks in this comparison, should be mounted directly on the CPU core, without the Athlon thermal transfer plate. This improves performance; the simulator test does not take this into account. Real-world performance might therefore be better.

Note that the tested heatsinks already represent a selection of some of the best available heatsinks; this means that even heatsinks that reached relatively high temperatures in this test might still be above-average when compared to cheaper heatsinks targeted at OEMs. All of the heatsinks tested would allow an Athlon-650 to run perfectly stable at 750MHz without overheating.


As you can see from the test results: once again, a Global WIN vs Alpha duel.

Note that when comparing Alpha P7125 and Global WIN VOS32, there is a discrepancy between the results obtained using the simulator and the results obtained in an actual computer. The reason is that the air can enter the Alpha heatsink only from one side - the side that faces the motherboard. Thus, the Alpha provides better performance when tested in "free space", since there is no motherboard next to the heatsink that could disturb the airflow. The VOS32, because of its different design, is not affected by this effect.

The Global WIN VOS32

Log in

Don't have an account? Sign up now