Category Archives: Uncategorized

Overclockers.net has an informative discussion going on their forums on the best way to apply thermal paste for the Intel i7 930.   Of course, the great advice there isn’t just for the i7, it could be applied to any processor using thermal paste also known as thermal grease.

There is a lot of good, real world advice in the discussion, but the one that caught our eye the most was this quote from user, “Windfire“, they said,

I think which the method is not as important as the skill of the person who does the application.

We couldn’t agree more!  Thermal paste or thermal grease as it is more commonly known, is the best thermal interface material available.  For overclockers, labs, prototyping and small volume shops it works in the application process.  But, where that breaks down is when you have a large production team at a major EMS.  The issues that we see here at ATS with teams we work with from Celestica, Flextronics and others is:

• inconsistent team training. So one team applies the thermal grease differently than another leading to poor thermal contact.
• Failure to consider that thermal paste/thermal grease has a shelf life.  We see this regularly when it’s not rotated and the result is poor thermal performance.

What to do about this?  Well, use Phase Change Material, it’s got very good thermal impedance without any of the drawbacks of grease.  One of those drawbacks is long term reliability.  Thermal grease often is overestimated in its reliability.  ATS has written a white paper on this topic that our readers are welcome to download here, entitled, “Long Term Thermal Grease Reliability”

As another resource for our readers, have a check on our March posting, “Thermal Interface Material Roundup“.  In that post we covered a number of thermal interface, thermal paste/thermal grease, and other providers.

The use of liquid metals as a cooling method is on the peripherery of high performance electronics thermal management.  The thermal conductivity of liquid metals can be orders of magnitude greater than water, glycol, or engineered dielectric fluids.   This means an equivilent amount of cooling can be performed using much less mass flow rate.  The requirements for pumping and pressure heads are reduced, which can in turn increase reliability.   To read more about this important technology, what’s needed to make it work, and how you might apply it in place of standard heat sinks for your thermal engineering and thermal design work, please click to our white paper article in QPedia, “Fundamentals of Magenetofluiddynamic Pumps“.