Socket 370/A Heatsink Review

---

OEM heatsinks are known not so much for their amazing performance but for their affordability and offers of decent cooling performance. Sort of a Ford Topaz line of heatsinks! They're inexpensive and they work... for the most part.

They work even better if you take a few moments to make one little modification. The modification is very simple and the improvement is substantial enough to make this decent heatsink cooler by an extra 6 degrees! More on that in a second though...

The Socket 370 heatsink Heatsink measures: 61mm x 66mm x 40 mm Fan measures: 60mm x 60mm x 15mm Fan 12V 1.4w ball bearing w/h RPM monitor Thermal pad on base Cost $12

The Synthetic Heat Test

To test out the socket 370/A cooler we used the FrostyTech Synthetic Test Apparatus. The device allows us to accurately load a heatsink with a known and consistently repeatable 50W heat load.

The heat sink itself is tested on two different copper die templates. Each template has an imbedded thermistor. Temperature values are recorded from this thermistor once they have stabilized during the test.

A small copper die template represents the surface area associated with FC-PGA or SocketA processors. FC-PGA type CPU's have only a small silicon core in contact with the base of the heatsink. The other copper die template is slightly larger and corresponds to the surface area of a Celeron PGA processor which has larger metallic heat spreader plate in contact with the heatsink.

Some heatsinks have thin bases which are poor at distributing heat from the processor across the base of the heatsink. The two individual tests serve as an important guage for heatsinks in both cooling situations.

FrostyTech Synthetic Test Temperature Results (120V 50Watts heat load)
	Mfg.	Model	Ambient Temp. (C)	Large block Rise Above Ambient(C)	large block temp (C)	small block rise above ambient (C)	small block temp (C)	Thermal compound	Thermal pad	No. Fans	Fan Noise	Clip
1	FrostyTech	CuSink	27.7	29.8	57.5	41.9	69.6	AOS 52029KY	-	1	low	stiff
2	KingCooler	Thermal buster	25.7	31.8	57.5	43.1	68.8	AOS 52029KY	yellow	1	none	stiff
3	Socket 370/A Cooler	OEM	27.8	36.8	64.6	42.2	70+	AOS 52029KY	therpad	1	low	stiff
4	ElanVital	c5010t12m	27.8	41.4	69.2	42.2	70+	AOS 52029kY	Lead	1	low	easy

The results of our testing with the socket 370/A cooler show nothing more than average performance with a 36 C rise in temperature above the abient. It faired well below that of our Frosty CopperSink and KingCooler's Thermal Buster reference heatsink. It only managed to perform slightly better than that of the the ElanVital Celeron PGA heatsink.

The results are not too impressive for the socket 370/A heatsink under high load from a small surface area FC-PGA type die.

Amazingly, this processor performance even poorer when we first tested it. The socket 370/A  comes out of the box with a small thermal pad stuck to the base. The thermal pad is meant to help transmit the heat from the cpu to the heatsink regardless of poor surface finish. As the socket 370/A has a brushed base, this would seem like a welcome feature.

In actuality, the pad proves to insulate the cpu more than it effectively transfers heat. We removed this little "thermal" pad and our temperature readings dropped by 6 degrees instantly!

Conclusions:

All in all, this socket 370/A heatsink looks good for its large size but really makes an average choice for FC-PGA or Socket A processors. It proves more than adequate when used for cooling older Celeron PGA-type processors however.

It may be able to keep a processor from melting under its own load but it definitely won't keep it anywhere near room temperature.