Here’s a little teaser for next week, as I have to work on other topics at the same time as all the graphics cards. And that’s where it gets exciting! In my current, detailed investigations into the laboratory analysis of processors, graphics processors and associated heatspreaders, I have been able to gain interesting insights that expose the long-standing myth that “a lot of thermal paste does no harm” as completely pointless. Using precise measurement techniques and taking into account empirical values and measurement data from industrial partners in the cooler and thermal paste industry, I want to demonstrate the extent to which the efficiency of the thermal transition is primarily determined by the quality of the thermal paste and the variable layer thickness, as a direct function of the surfaces. It has already been shown that the microscopic surface quality of the components and the associated irregularities play a much more important role than blindly relying on an excessive amount of paste or the tiresome discussion about the correct application method.
The measurements carried out – as well as the illustrations in the new database evaluation – prove that even slight variations in the applied paste layer lead to greatly differing thermal performances, even if an even distribution and adequate wetting of the surfaces is achieved. These findings refute the popular misconception that a generous application of thermal paste brings a significant advantage in the hope that what is too much will be squeezed out by itself. On the contrary, I found that an excessive amount not only offers no additional benefit, but in extreme cases can even lead to uneven layer formation, thickening and structural damage to the paste, resulting in increased bleeding (separation processes of the mixture) and thus to a deterioration in heat dissipation and durability.
In view of these findings, I intend to revise my database next week with regard to the search criteria adapted to the measurements and the evaluation of the behavior at very thin layer thicknesses. This update will provide a more practical reflection of real-life conditions and will also take into account the experience of overclockers and industrial manufacturing processes. For me, there is no question that it is almost irrelevant how the thermal compound is applied, as long as the final layer is as thin as possible and appropriately adapted to the specific surface properties without air inclusions. This finding underlines the importance of a differentiated approach that is adapted to the individual conditions of the hardware. I will also analyze different surfaces and prove that “flatter and smoother” does not necessarily offer more benefits.
The combination of my own measurement data with the extensive experience from the industry proves that a precise and controlled application to the respective environment is far more important than the simple application of large quantities of thermal paste. Scientific evidence shows that optimum thermal performance is primarily achieved by creating a thin, evenly distributed layer that is precisely adapted to the surface properties. This approach not only avoids wasting material, but also reduces the risk of mechanical problems that can occur if the layer is too thick.
I can only advise you to critically question conventional assumptions and blanket recommendations. The results obtained will help users to adopt a more flexible and well-founded approach to thermal pastes, taking into account both the individual circumstances of the hardware used and practical experience from various areas of application. With the upcoming update of my database, I will take a further step towards an even more precise and realistic evaluation of thermal behavior in order to provide end users with a sound basis for their decisions and thus contribute to sustainably stable and efficient thermal performance. After all, not every paste that shines with high thermal conductivity will ultimately be able to deliver what it promises at first glance.
So, and now you can continue to enjoy the weekend, I’m off to take another round of measurements…
26 Antworten
Kommentar
Lade neue Kommentare
Mitglied
1
Veteran
Veteran
Mitglied
Mitglied
Mitglied
Urgestein
Urgestein
Veteran
Veteran
Urgestein
Urgestein
Mitglied
Veteran
1
Neuling
Urgestein
1
Alle Kommentare lesen unter igor´sLAB Community →