Thermal Pad Roundup Tests Performance Under Compression

A new thermal pad roundup from Igor's Lab has tested ten pads and one thermal putty, revealing that the thermal conductivity figures printed on packaging are often misleading. The central finding is that practical performance depends on a material's behavior under compression, its effective heat transfer, and the residual force it exerts on a cooler assembly after installation.

Test Methodology and Key Findings

The thermal measurements were conducted using a Nanotest TIMA5 system, which operates on the principle of the ASTM D5470 standard. The test placed each material between temperature-controlled measurement bodies, recording heat flow, temperature, Bond Line Thickness (BLT), and mechanical load across several defined thickness stages. The reported effective thermal conductivity, λeff, describes the combined behavior of the material, its actual layer thickness, and its contact under test conditions, not an unchanging bulk property of the raw material.

For practical evaluation, the comparison used 69 kPa as a low-load marker, 172 kPa as the normal comparison range, and 345 kPa as an upper practical benchmark. Loads above 1,500 kPa were considered irrelevant for the application context. The test measured stabilized residual force after settling, not short-term insertion peaks. Values below approximately 1 N indicate very low remaining load, not installation without any force.

Performance at 1 mm and in Simulations

The source states that at approximately 1 mm BLT, results can differ substantially from a material's optimum performance. The test recorded the following effective thermal conductivities at this thickness:

A DRAM simulation produced minimum temperature increases for several materials. The Thermal Grizzly Minus Pad Extreme showed a 12.28 K increase, while the Aairhut 15 W/mK pad reached 14.08 K. The ARCTIC TP-4 and EC360 Platinum Soft pads resulted in increases of 15.06 K and 15.70 K, respectively. The EC360 Blue pad reached 34.56 K, and the HAVIRI set reached 41.36 K.

In a VRM simulation, the Minus Pad Extreme recorded the lowest individual temperature at 59.38 °C. The Aairhut 20 pad ranged as high as 63.71 °C, and the HAVIRI material reached 63.47 °C.

Assessment of Top Performers

The roundup identifies EC360 Platinum Soft as the thermal performance leader, achieving 11.84 W/mK. However, its optimum requires 327.6 kPa of stabilized residual load, which is close to the 345 kPa practical benchmark. The source notes this makes its maximum value "less attractive for shared GPU, memory and VRM coldplates than the headline result suggests."

Thermal Grizzly Minus Pad Extreme is described as "the most balanced premium candidate in the tested field." It combines 9.78 W/mK effective conductivity with a broad high-performance window of approximately 45% and almost no residual force after settling. For lower-cost options, the ARCTIC TP-4 and Aairhut 15 W/mK pads provide a balance of effective conductivity, low force, and installation tolerance.

The TG Putty Advance offers lower mid-range thermal performance but is highlighted as useful where different component heights make fixed pad thicknesses difficult to manage.

Practical Implications for Builders

The source emphasizes that a thermal pad does not operate according to the number printed on its packaging but according to its actually achieved thickness, its adaptation to surfaces, and the permanent force it exerts. A nominally 1 mm-thick pad does not automatically remain 1 mm thick during operation. Some materials reach their thermal optimum at around 800 µm, while others need compression to approximately 700 or 650 µm.

The author states, If a supplier prints 20 W/mK and I measure 5.7 W/mK λeff, this is not automatically proof that the manufacturer’s specification is incorrect. However, the test found that when two products from the same supplier were compared, the nominal 20 W/mK pad performed worse than its 15 W/mK variant, raising questions about the significance of such packaging figures.

The roundup does not include material analysis or detailed microscopy, focusing instead on the immediately relevant user question of how a complete pad behaves under real compression. The prices cited are snapshots from September 13, 2026, and refer to the tested size where it could be unambiguously identified, excluding shipping costs.