The necessity of heatspreaders on system memory has been debated countless times on just about every tech forum online.  This article takes a look at the composition of an OCZ Technology XTC heatspreader installed on a dual channel DDR2 kit to see if it really does anything useful.  The memory in question is a 1GB kit of OCZ Technology's "DDR2 PC2-5400 Gold GX XTC Dual Channel" [product link].  Not the latest and greatest in their product line up, but it is still decent stuff for anyone looking for DDR2 667MHz memory.

The images below provide a look at the front and back of the OCZ Technology Gold kit in its original form.  The XTC (Xtreme Thermal Convection) heatspreader with the stylized "Z" makes these modules easy to identify as an OCZ Technology product.  While I have always been somewhat suspect of the need for heatspreaders on many types of memory, these raised an eyebrow for me even more so than the typical design.

Looking at the assembled modules shows that there is a thin strip of perforated metal affixed to the memory chips with what appears to be common double sided tape.  Looking on the edge of the modules reveals that this perforated strip is in turn bonded to the metal that goes around the edge, perhaps there just to complete the look and provide some extra mass.  While the modules look cool, the focus of this article is to see if they actually doing anything to keep the chips cool.

So begins the process of disassembling the memory modules.  The images below detail the first step, removing the trim metal that is bonded to the perforated strip.  This was extremely easy, and by just getting a finger nail under one corner, the golden frames can be pryed up from either side in seconds.

I'm not a materials engineer, but the adhesive used to bond the trim to the perforated metal doesn't seem to have any special properties to aid in thermal conductivity.  It seems like if anything, it would be an insulator, except for those area where there are voids in the adhesive film and potential metal on metal contact occurs.

The same can be said for the bond between the perforated metal and the memory chips.  It looks like typical double sided tape, and while I can't analyze its composition, it really seems more like an insulator than a conductor.  On top of my theory that the adhesive may be an insulator, the bulk of the chip's surface area is not bonded to metal.  The wide open nature of the perforation would result in most of the chip's top surface being exposed to air if it weren't for the adhesive.