You probably already know that an all-in-one (AIO) liquid cooler is a self-contained unit — a pump, radiator, and cold plate (the metal block that presses against your CPU) all plumbed together and ready to bolt in. What the box usually doesn’t tell you is that there’s a thin layer of thermal paste — the heat-conducting compound between your CPU’s heat spreader and that cold plate — pre-applied at the factory. Thermal paste is what bridges microscopic surface imperfections so heat flows efficiently from processor to cooler. When it’s good, you’ll never think about it. When it’s dried out, cracked, or uneven, your CPU can run 10–20 °C hotter than the cooler’s design allows — and you’ll spend weeks chasing a bottleneck that a $10 tube and 20 minutes of your time could have eliminated on day one.
This guide is the case for making repasting a universal first step whenever you unbox an AIO — not because your cooler is defective, but because the supply chain, shelf time, and factory application methods mean that “good enough out of the box” is genuinely a coin flip. We’ll walk through how to diagnose the problem, how to do the swap correctly, and which compounds the community consistently reaches for on high-TDP builds.
The Supply Chain Problem Nobody Talks About
Here’s the uncomfortable truth: thermal paste has a shelf life, and AIOs sit in distribution warehouses for weeks or months before they reach you. Factory application lines apply paste in bulk, often using automated stencils or dot-dispensers calibrated for average cold plate geometry — not the specific unit in your box. By the time a cooler has sat on a pallet, shipped across an ocean, cleared a distribution center, and landed at your door, that factory compound may be well past its prime.
The pattern across owner feedback is striking and consistent. Reviewers at Tom’s Hardware and aggregated owner reports for the Arctic Liquid Freezer III Pro document multiple independent users receiving units with preapplied paste that had visibly dried out — brittle around the edges, cracked in the center, or simply chalky in texture rather than viscous. Those owners self-reported idle CPU temperatures in the 40–50 °C range on coolers rated to handle processors in the 150–200 W TDP envelope. After cleaning the cold plate and applying fresh aftermarket compound, those same owners self-reported drops to the 29 °C range at idle — a nearly 20 °C improvement attributed entirely to paste condition, not any change to pump speed, fan curve, or mounting pressure.
Corsair Nautilus owners echo this independently. KitGuru’s coverage of owner-reported Nautilus results includes a documented 10 °C improvement after swapping factory paste, with no other variables changed. And critically: the Nautilus sits at a premium price tier. The notion that this is a budget-cooler problem doesn’t hold up.
A secondary factor worth naming explicitly: the Arctic Liquid Freezer III Pro also carries recurring owner complaints about mounting system refinement — specifically, that the bracket system can allow slight rocking of the cold plate under clamping force, which creates uneven contact even when paste condition is fine. This is a compounding issue: degraded paste plus uneven contact means the factory numbers on the box may be structurally unachievable without user intervention. igorslab.de’s long-form cold plate contact analysis confirms that mounting pressure consistency is a material variable in AIO thermal performance, not a minor footnote.
By the Numbers
| Scenario | Self-Reported Idle Temp (°C) | Source |
|---|---|---|
| Arctic LF III Pro — factory paste (dried) | 40–50 °C | Owner self-reports via Tom’s Hardware |
| Arctic LF III Pro — aftermarket repaste | ~29 °C | Same owners, post-repaste |
| Corsair Nautilus — factory paste | Baseline | Owner reports via KitGuru |
| Corsair Nautilus — aftermarket repaste | ~10 °C lower | KitGuru owner compilation |
All temperature figures are owner self-reported under varying system configurations. Individual results depend on CPU TDP, ambient temperature, fan curves, and mounting torque.
Kryonaut vs. Duronaut: The Decision Frame for High-Core-Count Builds
If you’re repasting, you’re going to buy a compound, and the Kryonaut vs. Duronaut question comes up constantly on 360mm AIO builds paired with high-core-count CPUs — the i9-14900K, the Ryzen 9 9950X, or anything in the Threadripper family where sustained AVX workloads push cold plates into territory where paste performance genuinely separates compounds.
Thermal Grizzly Kryonaut is the incumbent. TechPowerUp’s thermal compound comparatives consistently place it in the top tier of non-conductive pastes, with published thermal resistance values that make it competitive within a degree or two of the best alternatives. It spreads easily, seats quickly under mounting pressure, and behaves predictably across a wide cold plate area. The tradeoff: Kryonaut has a documented pump-out concern under sustained high-temperature cycling — the compound can gradually migrate away from the die center over months of repeated thermal expansion and contraction. For intermittent gaming workloads this matters less; for a 3D rendering farm running 18-hour sustained loads, it matters more.
Duronaut — also from Thermal Grizzly — is specifically formulated to resist that pump-out behavior. It’s stiffer at room temperature, which means application technique matters more (more on that below), but owners paired with the Arctic Liquid Freezer III Pro during AVX stress testing — including one reviewer specifically calling out Duronaut by name in their repaste write-up — report stable long-run thermal performance that Kryonaut struggles to match under continuous sustained load.
The decision rule: If your workload is gaming or mixed-use with thermal cycles that return to idle regularly, Kryonaut is easier to apply and delivers excellent results. If you’re running sustained all-core workloads — content rendering, scientific compute, Prime95 for extended periods — Duronaut’s pump-out resistance makes it the better long-run investment despite the slightly higher application friction.
Arctic MX-6 deserves mention as the value-tier alternative. It won’t match either Grizzly compound at the margin, but across aggregated owner reviews it consistently outperforms every factory preapplication we’ve seen documented, and it’s non-electrically-conductive, non-corrosive, and straightforward to work with. For a $25 Thermalright Assassin Spirit, it’s arguably the right call.
How to Actually Do the Repaste: Cold Plate to Cure
Step 1 — Degrease the cold plate. Copper cold plates scratch more easily than you’d expect. Use 99% isopropyl alcohol (IPA) on a lint-free cloth or a high-quality microfiber. Do not use paper towels — the fiber texture is abrasive enough to leave micro-scratches on polished copper surfaces. Apply gentle, circular pressure. If old paste is caked on, let a IPA-saturated cloth sit on the surface for 30–60 seconds to soften it before wiping. igorslab.de’s cold plate surface analysis notes that even minor surface scratches can increase thermal resistance by creating air pockets under paste, so this step is worth doing slowly.
Step 2 — Clean the CPU heat spreader. Same process: 99% IPA, lint-free cloth, circular motion. The IHS (integrated heat spreader — the metal lid on top of your CPU) should be completely clean and dry before any paste touches it.
Step 3 — Apply the paste. For AIO cold plates with a circular or near-circular contact area, a single centered pea-sized dot (roughly 4–5mm diameter) is the standard recommendation and what most compound manufacturers specify. The mounting pressure of the cold plate will spread the paste to cover the die area. Spreading paste manually with a card or spatula before mounting is unnecessary for most compound viscosities — and with a compound like Duronaut, manual spreading may actually trap air bubbles. The exception: very large cold plates designed for LGA4189 or TR5 (Threadripper 5000) footprints, where a thin cross or small X pattern ensures full die coverage given the die cluster spread.
Step 4 — Mount carefully and check torque. For AIOs with Intel LGA1851 or AMD AM5 retention hardware, tighten mounting screws in a diagonal pattern (corner to opposite corner) rather than sequentially around the perimeter. This ensures even cold plate contact. HardwareLuxx’s mounting pressure analysis found that uneven clamping force can produce contact variance of several degrees across the same paste application — the screw pattern is not a minor detail.
Step 5 — Run a short load cycle immediately. A single Cinebench R23 nT run is enough to seat the paste under thermal cycling. Check temperatures after this first run and again after 24 hours of normal use. Most compounds reach close to their rated performance within the first few thermal cycles.
Frequently Asked Questions
How can I tell if my AIO’s preapplied paste is dried out before I even install it? Remove the cold plate protective cover (usually a thin plastic film) and inspect the paste under direct light. Fresh factory paste is typically gray or silver, slightly glossy, and visibly pliable — it should look wet. Dried paste will appear matte, may show cracking or a chalky surface texture, and in some cases will have pulled away from the edges of the application area. If it looks like anything other than a uniform, slightly reflective layer, treat it as compromised and repaste before installing.
Will repasting void my AIO warranty? This depends on the manufacturer’s warranty terms and your region’s consumer protection laws. Most major AIO manufacturers — Arctic, Corsair, Noctua, NZXT, ASUS — do not explicitly void warranties for repasting, because the cold plate is an external service surface. However, disassembling internal pump or radiator components will typically void coverage. Read your specific warranty documentation before proceeding, and photograph the factory paste condition before cleaning it off — that documentation can matter in a warranty claim scenario.
Is Kryonaut or Duronaut better for a 360mm AIO paired with a high-core-count CPU? For sustained all-core workloads — rendering, scientific compute, extended Prime95 — Duronaut’s pump-out resistance makes it the better long-run compound. For gaming or mixed workloads with regular idle returns, Kryonaut is easier to apply and performs within margin. TechPowerUp’s compound benchmarks support this framing. The gap between them at peak single-run performance is small; the gap in six-month performance under thermal cycling is more meaningful.
How do I clean a copper cold plate without scratching it? Use 99% isopropyl alcohol and a lint-free microfiber cloth. Avoid paper towels, cotton swabs with abrasive tips, and any mechanical scraping tool. Apply the IPA, allow it to soften dried paste for 30–60 seconds, then wipe with gentle circular pressure. Copper is soft — it polishes and scratches more easily than nickel-plated surfaces. Patience matters here.
Should I use a pea-sized dot or spread the paste on an AIO cold plate? A centered pea-sized dot is correct for standard AIO cold plates. Mounting pressure under a properly torqued bracket spreads it uniformly. Manual spreading risks trapping air and adds no benefit for circular cold plates. The exception is very large multi-chiplet die coverage (TR5, HEDT platforms) where a cross or X pattern ensures the paste reaches all active die clusters before the cooler seats.
How often should I repaste an AIO cooler after the initial application? Most premium compounds — Kryonaut, Duronaut, MX-6 — are stable for 2–3 years under normal gaming or workstation workloads. If you’re running sustained AVX or rendering loads daily, revisit at the 12–18 month mark. A reliable signal that it’s time: idle and load temperatures have risen 3–5 °C with no other system changes. igorslab.de’s long-term degradation analysis confirms that compound pump-out and oil separation are the primary failure modes, and both are gradual and temperature-measurable before they become severe.
The Bottom Line
The factory paste on your AIO is a starting point, not a guarantee. The supply chain doesn’t promise freshness, the automated application line doesn’t know your cold plate’s specific surface finish, and the premium price on a Corsair Titan 360 RX LCD or NZXT Kraken Elite does not exempt you from this reality. The owner-reported evidence — nearly 20 °C recoverable on the Arctic Liquid Freezer III Pro, 10 °C on the Corsair Nautilus — puts the cost-benefit math firmly on the side of repasting.
If your workload is gaming or mixed-use: repaste with Kryonaut or Arctic MX-6, use the pea-dot method, follow the diagonal torque sequence, and revisit in two to three years.
If you’re running sustained all-core loads — rendering farms, AVX workloads, overclocked Threadripper builds: Duronaut is the compound to reach for, and the 12–18 month re-check should be on your maintenance calendar.
Twenty minutes and the cost of a tube of compound is a small investment against the thermal headroom you paid for when you chose a premium AIO. Do it before the first boot, and your temperature readings will actually reflect the cooler’s design — not the shelf life of a factory application.