AIO Cooler Thermal Paste: Why the Pre-Applied Stuff Often Fails
If you’ve just unboxed an all-in-one liquid cooler — an AIO, which is a self-contained pump-and-radiator unit that mounts directly to your CPU — you’ve probably noticed a thin grey smear pre-applied to the copper or nickel cold plate (the flat metal surface that touches your processor). That smear is thermal paste, a heat-conducting compound that fills the microscopic air gaps between the cold plate and your CPU’s heat spreader. Air is a terrible conductor of heat; paste is much better. The whole point is to let heat escape your processor as fast as possible. When the paste works correctly, your CPU runs cooler and quieter. When it doesn’t — and the factory-applied stuff frequently doesn’t — you can end up with temperatures 10 to 20 degrees Celsius higher than the cooler is actually capable of, fans screaming at full speed, and a nagging suspicion that you bought the wrong cooler. This guide explains why pre-applied paste fails, how to identify the problem, and exactly what to do about it.
Why Pre-Applied Paste Is a Gamble, Not a Guarantee
The core issue is the supply chain. Thermal paste has a working life. Most quality compounds stay viable for 2–3 years sealed, but once the protective film is removed or the paste is exposed to repeated heat cycles in a warehouse or during shipping, the carrier oils can separate and the compound begins to dry out. By the time a unit reaches your doorstep — especially if it sat in a fulfillment center through a summer — the paste on the cold plate may already be compromised.
Multiple owners of the Arctic Liquid Freezer III Pro 360 (both the black and white variants) have documented exactly this scenario in aggregated review data. One owner reported idle CPU temperatures of 40–50°C and sustained fan noise before repasting; after cleaning the cold plate and applying fresh compound, idle temps dropped to 29°C and the system became near-silent under the same workload. That is not a marginal difference — that is the difference between a well-functioning cooler and one that behaves as if it were two size classes smaller. Reviewers at Tom’s Hardware have noted that the Liquid Freezer III series is capable of elite-tier performance, which makes the dried-paste failure mode particularly frustrating: the hardware is fine, but the interface material is letting it down.
A Corsair Nautilus 360 RS owner documented a similar experience, explicitly recommending removing pre-applied paste and replacing it with quality compound and reporting a full 10°C reduction in peak load temperatures. Across aggregated long-term reliability reports at KitGuru, the pattern repeats: pre-applied paste underperforms fresh compound, and the gap widens as the paste ages.
This is not a brand-specific defect. It is a structural problem with the category. Any AIO that ships with pre-applied paste is subject to the same supply-chain uncertainty. The paste is applied at the factory; the cooler may ship weeks or months later; you may install it months after that. By the time the cold plate touches your CPU, you have no visibility into how many heat cycles the paste has seen or whether the carrier oils have separated.
The Numbers: What Repasting Actually Delivers
To understand why this matters, it helps to look at what the compound itself contributes — independent of the cooler’s overall design.
By the numbers:
| Scenario | Reported Peak Temp (Ryzen 9 9900X, sustained AVX) |
|---|---|
| Arctic Liquid Freezer III Pro 360 — stock paste, dried | ~48–52°C idle / fans audible at rest |
| Arctic Liquid Freezer III Pro 360 — repasted, fresh MX-6 or equivalent | ~29°C idle / near-silent |
| Thermal Grizzly Duronaut pairing (same cooler class) | ~62°C under sustained 24-thread AVX load |
The Duronaut figure, reported by owners pairing it with coolers in this performance tier and cited in TechPowerUp’s compound roundup data, represents a well-optimized baseline under one of the more brutal sustained workloads you can run. The dried-paste scenario produces results substantially worse than that — not because the cooler is bad, but because the thermal interface is failing.
Igor’s Lab’s thermal paste longevity analysis confirms the physics: as carrier oils separate from the solid particle matrix, thermal conductivity drops and contact resistance increases. The CPU junction temperature rises, the fan curve responds by ramping fans, and the user concludes the cooler is undersized. In many cases, the cooler is fine.
How to Spot Dried Paste Before Installation
You can do a quick visual and tactile check the moment you remove the cold plate’s protective cover:
Color and texture. Fresh thermal paste should be uniformly grey (or white, depending on compound), with a smooth, slightly glossy surface. Dried paste looks chalky, may have cracked edges, or shows visible separation — a darker center ring with lighter, dried-out borders.
Coverage pattern. Some manufacturers apply a thin pre-spread layer; others apply a center dot. Either is fine when fresh. If the paste looks uneven, has air pockets visible in the layer, or has pulled away from the center of the cold plate, treat it as suspect.
The wipe test. Run a clean fingertip (or a gloved fingertip if you prefer) lightly across the edge of the paste. Fresh compound feels smooth and slightly tacky — it resists being wiped cleanly because it’s still viscous. Dried compound flakes or crumbles, or wipes off in dry chunks with no resistance.
If any of these signs are present, do not install the cooler. Clean the cold plate with isopropyl alcohol (90% or higher concentration) and apply fresh compound. You are not voiding your warranty in any meaningful way — more on that in the FAQ below — and you are almost certainly improving your final temperatures.
Choosing a Replacement Compound: MX-4, MX-6, NT-H2, Kryonaut, and Duronaut
Not all thermal paste is equal, but the gap between mid-tier and flagship compounds is smaller than marketing suggests. Here is how the field breaks down for AIO repasting specifically.
Arctic MX-4 and MX-6 are the workhorse choices for this use case. Both are electrically non-conductive — meaning a slight misapplication will not short your motherboard — and owners consistently describe application as low-stress. MX-6 reviewers note the spreading consistency is stiffer than older Arctic Silver formulations, which makes it slightly less forgiving for beginners but more stable under high-heat cycling. Either compound is a meaningful upgrade over degraded factory paste, and neither requires the careful surface-prep obsessiveness that metal-particle compounds demand.
Noctua NT-H2 sits a step above in measured thermal conductivity. Reviewers at HardwareLuxx specifically call out the included cleaning wipes as a practical bonus for exactly the repasting use case — you get the compound and the prep material in one package, which matters when you are doing this mid-build with a processor already in the socket. If you are repasting an existing install rather than a fresh build, NT-H2’s bundled wipes are genuinely useful.
Thermal Grizzly Kryonaut is the benchmark compound in most published roundups at TechPowerUp. It performs at the top of the non-metallic tier, it works on every socket, and it is the reference point against which other compounds are measured. The tradeoff is price and application — it is less forgiving if you over-apply, as excess paste near socket edges requires careful cleanup.
Thermal Grizzly Duronaut is the newer, higher-viscosity formulation aimed at sustained high-TDP workloads. Owners pairing it with 360mm AIOs on Ryzen 9000-series CPUs report the approximately 62°C sustained AVX figure cited above, which is competitive for that workload tier. Duronaut’s higher price point is justified if you are running a Threadripper or a Core i9 under continuous professional load; for a mainstream gaming build, Kryonaut or NT-H2 is sufficient and costs less.
The honest decision rule: If your budget is under $10, get MX-6. If you want the NT-H2 wipes for a repaste-in-place job, spend the extra few dollars for NT-H2. If you are running sustained professional compute workloads on a 250W+ CPU, Duronaut earns its premium. The performance difference between mid-tier and flagship compounds is real but measured in single-digit degrees Celsius under most consumer workloads — the bigger gain is simply replacing degraded paste with anything fresh.
Application Method for an AIO Cold Plate
Cold plates on AIOs differ from tower air cooler heat pipes in one important way: the contact surface is typically larger and flatter, which affects spread method.
Center dot is the most universally safe method. A pea-sized or rice-grain-sized dot in the geometric center of the cold plate will spread under mounting pressure into a thin, even layer covering most of the IHS (integrated heat spreader) surface. It works reliably across Intel LGA1700/LGA1851 and AMD AM5 socket geometries.
X pattern adds a dot at each corner of a light central X, which helps pre-distribute compound for larger IHS surfaces like Threadripper or EPYC. For standard desktop CPUs it is unnecessary and can encourage over-application.
Pre-spread (thin layer) is what manufacturers attempt at the factory. Done correctly with fresh compound, it is actually the most consistent method — you eliminate the spread uncertainty entirely. Done incorrectly with too much compound, it creates squeeze-out that is difficult to clean. If you pre-spread, aim for a layer thin enough that you can barely see the cold plate surface through it.
For most AM5 and LGA1851 builds, a center dot is the right call. It is fast, it is consistent, and it leaves almost no cleanup risk.
Frequently Asked Questions
How can I tell if my AIO’s pre-applied paste has dried out before I even install it? Remove the cold plate’s protective film or cover and inspect the paste under good lighting. Fresh paste is uniformly smooth and slightly reflective. Dried paste looks chalky, shows cracking at the edges, or has visible separation rings. A light fingertip touch at the edge will confirm: fresh paste resists cleanly; dried paste crumbles or flakes. If you see any of those signs, clean and repaste before mounting.
Does repasting void my AIO warranty? In practice, almost no manufacturer can determine whether you repasted, and most warranty claims involve pump failure or leak failure — not thermal performance disputes. Technically, some warranty language covers “unmodified product,” but repasting the cold plate surface is universally treated as user maintenance, not modification. No credible service record in aggregated community experience at KitGuru or Tom’s Hardware documents a warranty denial attributable to repasting.
What is the best application method — dot, X, or spread — for an AIO cold plate specifically? Center dot for standard desktop CPUs (AM5, LGA1851, LGA1700). It self-spreads under mounting pressure into an even layer. Pre-spread works well if you use a plastic card or spatula and keep the layer thin. X pattern is useful for larger IHS surfaces (Threadripper, HEDT) but unnecessary for mainstream desktop sockets.
How often should I repaste an AIO cooler over its lifetime? Igor’s Lab’s longevity analysis suggests most quality compounds maintain performance for 3–5 years under normal cycling. A practical schedule: repaste at first installation if the factory paste shows any degradation signs, then again at 2–3 years if you are monitoring temperatures and notice a gradual upward drift. A 5°C unexplained rise in your normal load temperatures, with no changes to airflow or ambient, is a reliable signal to repaste.
Is the paste quality difference between MX-4, MX-6, Kryonaut, and Duronaut meaningful enough to justify the price gap? Between MX-4 and MX-6: minor real-world difference, MX-6 is the current formulation and worth the small price delta. Between MX-6 and Kryonaut: measurable in benchmarks (1–3°C on most desktop CPUs), arguably worth it if you are pushing TDP limits. Between Kryonaut and Duronaut: Duronaut earns its premium specifically for sustained high-TDP workloads above ~200W — for standard gaming builds, Kryonaut is sufficient and costs less. The single biggest thermal gain, by a wide margin, is replacing degraded factory paste with any of these compounds. Arguing over Kryonaut versus MX-6 when running dried factory compound is optimizing the wrong variable.
The takeaway is simple: an AIO cooler is a precision thermal system, and the one component you have no quality control over at purchase time is the paste on the cold plate. Thirty minutes and a quality compound can recover the full performance the cooler was designed to deliver. If your cooler arrived with paste that looks even slightly suspect, replace it before you close the case. The benchmarks are clear on this — and the fix costs less than most people spend on cable combs.