CPU Thermal Paste Application: Methods, Quantities, and What Actually Matters
The internet hosts vigorous debates about thermal paste application patterns. The data shows that the method matters far less than getting coverage right, using the correct quantity, and cleaning the prior installation properly. Here is what actually moves the temperature needle.
Thermal paste fills microscopic gaps between a CPU heat spreader and a cooler contact plate. Both surfaces appear flat but contain irregularities at the microscopic level that trap air when pressed together. Air is an effective thermal insulator. Thermal paste, with conductivity typically in the range of 1 to 12 W/mK depending on formulation, replaces that trapped air with something far less insulating. The goal is not to apply paste so thick it conducts heat on its own, but to apply enough that no air pocket survives between CPU and cooler once the mounting pressure is applied.
Application method comparison
Several patterns are commonly recommended: a pea-sized center dot, an X shape, a thin pre-spread layer, and a rice-grain bead. Extensive third-party testing from hardware reviewers including Igor's Lab and GamersNexus consistently shows the temperature difference between these methods, when done correctly and with the same paste quantity, is typically under 2 degrees Celsius. The cooler's mounting pressure and the spreader contact surface area drive the final distribution far more than the starting pattern.
| Method | Coverage behavior | Over-application risk |
|---|---|---|
| Center dot (pea-sized) | Spreads from center under mounting pressure | Low if quantity is correct |
| X pattern | Faster initial coverage at contact | Slightly higher |
| Full pre-spread | Pre-distributed, no spread needed | Highest, risk of trapped bubbles |
| Line bead | Useful for long rectangular HEDT IHS | Low, platform-specific |
For standard square IHS CPUs on Intel LGA and AMD AM5 platforms, the center dot is the least error-prone method because it produces a controlled spread radius proportional to mounting pressure rather than relying on pre-distribution accuracy.
Quantity matters more than pattern
Excess paste that migrates beyond the IHS edge can reach capacitors on the PCB. For non-conductive polymer compounds this is cosmetically ugly but electrically harmless. For metallic pastes containing silver, bridging between surface-mount components is a genuine short-circuit risk. The correct quantity for a standard desktop CPU is a ball roughly 4 mm in diameter, comparable to a small pea. After mounting, the paste should fill the contact area without significant overflow at all four edges.
Paste formulation types
| Category | Example compounds | Conductivity (W/mK) | Notes |
|---|---|---|---|
| Polymer-based | Noctua NT-H1, Thermal Grizzly Hydronaut | 8.5–11.8 | Non-conductive, safe for PCB contact |
| Carbon-based | Arctic MX-6, Cooler Master MasterGel Pro | 8.5–12 | Non-conductive, good longevity |
| Metal-based liquid | Thermal Grizzly Conductonaut | ~73 | Electrically conductive, delidded use only |
For a standard cooler-to-IHS installation, polymer or carbon paste in the 8 to 12 W/mK range is the correct choice. The temperature difference between a 12 W/mK polymer paste and a 73 W/mK liquid metal compound on a standard desktop CPU with IHS is typically 3 to 6 degrees Celsius under load. Liquid metal delivers its advantage primarily on delidded CPUs where it contacts the silicon die directly, not through the spreader.
The cleaning step is the most impactful step
If you are replacing paste on an existing build, cleaning thoroughness has a larger impact on results than paste selection. Dried polymer paste develops micro-cracks and loses contact conformity. Remove old paste completely with 90 percent or higher isopropyl alcohol on a lint-free cloth. Do not use acetone on plastic cooler shrouds or AIO pump heads. Allow surfaces to fully dry for 60 seconds before applying fresh paste. Any residue left behind from the prior application prevents the new paste from seating correctly against the contact surface.
Mounting sequence
Apply paste, position the cooler, and lower it straight down without sliding. Sliding the cooler after initial contact displaces paste unevenly and can introduce air pockets. Tighten mounting screws in a diagonal cross pattern, applying incremental pressure to all corners rather than fully seating one corner at a time. Uneven pressure on a rigid cooler base produces a rocking effect where one corner contacts more firmly than the opposite, leaving a thin paste distribution on one side and a thick one on the other.
When to repaste
For most polymer compounds, performance holds within acceptable range for two to four years under normal operating conditions. The practical trigger for repasting is a CPU temperature that has climbed 10 or more degrees Celsius compared to a logged baseline from when the paste was fresh, with no other explanation such as a case dust buildup or cooler fan degradation. Environments with large seasonal temperature swings accelerate compound degradation relative to a stable climate-controlled room.