Acoustic Samples: Listen to this Heatsink!

The fan really doesn't draw that much power or produce that much noise so this waveform image is slightly overstated I'd say. Sound is a difficult thing to image so your best bet is to always just listen to the heatsink by clicking on the MP3 file (click on the headphones).

FrostyTech Acoustic Sampling Chamber
Standard Waveform view of a 10 second recording. Click on the headphones to listen to an MP3 recording of this heatsink in operation.

Sound Level Measurements:

Listening to the actual noise a heatsink makes allows you to hear what pure numbers alone cannot get across. To further emphasize FrostyTech's reliance on cold hard facts when evaluating a heatsink, we also take Decibel readings with a sound level meter. These results may be higher than the manufacturer's listed specs, but then again these are real world measurements.

FrostyTech Sound Level Results
-	Manufacturer	Model	Noise Level	Scale
	Zalman	CNPS6500-AlCu	47.4/28.9 dB	quiet
	Intel	Socket 478	42.5 dB
	JMC	400011	48.0 dB
	Cho-Liang	CL-CB00881	50.2 dB
	AVC	117140	56.1 dB	louder

The JMC fan used on the 400011 comes in just behind the Intel heatsink which is not really that surprising.

Synthetic Temperature Test Results: *Ranked according to 100W interface die test results.

FrostyTech Synthetic Temperature Test Platform Results
	Mfg.	Model	Ambient Temp.	27mm large copper block		27mm large copper block	Thermal pad	No. Fans	Fan Noise	Clip force
				50W		100W
1	AVC	117140	22.6	37.8		50.4	none	1	med	med
2	Arkua	8568	22.4	37.8		51.7	none	1	med	med
3	Zalman	CNPS6500-AlCu	23.1	40.3		57.4	none	1	low	easy
4	Vantec	GSN-7025	22.5	41.7		58.5	none	1	med	med
5	Chauh Choung	DTA-478	22.4	42.1		59.0	carbon	1	low	med
6	Intel	Socket 478	22.4	41.8		59.8	carbon	1	low	med
7	Titan	TTC-MT3AB	22.8	41.5		61.2	none	2	loud	stiff
8	Dynatron	DC1206BM-S	22.4	43.2		62.0	none	1	med	med
9	AVC	117160	22.4	42.7		62.2	white pad	1	med	med
10	JMC	400011	23.0	43.4		63.6	pink	1	low	easy
11	Cho-Liang	CL-CB00881	23.6	41.2		65.0	none	1	med	easy
12	Vantec	GSN-7015	22.4	47.5		69.0	none	1	loud	stiff
				50W		100W

When the heatsinks are tested under 100W of heat you can begin to see where the different coolers are separated. In the case of the JMC 400011, it falls behind that of the stock Intel heatsink and this really shows that the design is outdated.

After all this is a heatsink which was originally intended to mount upon a socket 423 processor. Even though the JMC heatsink is rated to 2.0GHz, you could do better with stick Intel solutions so the incentive to switch is just not there.

Rise Above Ambient Temp. (c)
	Mfg.	Model	27mm large copper block		27mm large copper block
			50W		100W
1	AVC	117140	15.2		27.8
2	Arkua	8568	15.4		29.3
3	Zalman	CNPS6500-AlCu	17.2		34.3
4	Vantec	GSN-7025	19.2		36.0
5	chauh Choung	DTA-478	19.7		36.6
6	Intel	Socket 478	19.4		37.4
7	Titan	TTC-MT3AB	18.7		38.4
8	Dynatron	DC1206BM-S	20.8		39.6
9	AVC	117160	20.3		39.8
10	JMC	400011	20.4		40.6
11	Cho-Liang	CL-CB00881	18.2		41.4
12	Vantec	GSN-7105	25.1		46.6
			50W		100W

Well the JMC heatsink fairs better than both the Cho-Liang and the Vantec GSN-7015 in the rise above ambient results. In actuality, the 400011 is not as bad as we may have first thought.

It is not a great heatsink, but falls a short 3 degrees behind that of the Intel heatsink. A simple fan switch could be all that is needed to put the Arctic Ice back on track and at par with the stock Intel platform. This is a bit surprising because we had really expect the JMC heatsink to perform pretty poorly.

Just goes to show you, the test results are more important than any impression. So while the JMC Arctic Ice socket 423/478 (with right HSRM) is a sub-stock heatsink performer, all is not completely lost thermally.