How-To Geek’s Sydney Butler makes the case for treating liquid metal as a last-step enthusiast upgrade rather than routine maintenance. The evidence supports that caution, with an important qualification: impressive benchmark results and factory adoption do not automatically translate into a safe, worthwhile DIY modification.
The temperature gains are real—but context matters
One frequently cited example comes from Gamers Nexus’s September 2017 testing of Intel’s Core i9-7900X. With frequency fixed at 3.6GHz and voltage at 1.15V, its Blender workload showed approximately a 10°C improvement after a liquid-metal modification. A Prime95 comparison at those settings measured approximately 80°C before the modification and 68°C afterward.
Those are meaningful differences. They also came from delidding the processor and replacing the thermal interface beneath its integrated heat spreader, not simply changing the compound between an intact desktop CPU and its cooler. Gamers Nexus also adjusted the heat-spreader adhesive during the procedure. This was a processor modification, not a conventional repaste.
That distinction matters when deciding whether the risk makes sense. The results establish that improving that particular internal interface helped that particular processor. They do not establish that applying liquid metal above an intact heat spreader will shave 10°C off another PC.
The takeaway: a benchmark is evidence for its tested configuration, not a temperature coupon redeemable on every CPU.
Electrical conductivity leaves less room for error
ASUS explicitly warns that liquid metal contacting motherboard electrical components can cause a short circuit. Its support guidance discourages users from disassembling the thermal module or replacing its cooling materials themselves because doing so can affect cooling performance or cause damage.
That is the central difference from a compound specifically designed to be electrically non-conductive: keeping the material within the intended contact area becomes an electrical-safety requirement, not merely a matter of neatness. Thermal Grizzly labels Conductonaut as electrically conductive; the product’s thermal benefits do not cancel that property.
This does not mean every spill inevitably destroys an entire computer. ASUS describes a short-circuit risk, not a guaranteed outcome. Nevertheless, the possibility of hardware damage is enough to make casual experimentation a poor fit for a machine you cannot afford to lose.
Aluminum is a separate—and serious—problem
Even perfect containment cannot rescue an incompatible installation. Thermal Grizzly says Conductonaut must not be used with aluminum coolers and recommends nickel-plated copper bases for the best long-term performance. Its compatibility guidance is more specific than simply “use something metallic.”
Gamers Nexus demonstrated the consequences in a separate September 2018 experiment. After Conductonaut remained against an aluminum coldplate overnight, the surface became blackened and chalky, and the aluminum became brittle. That documents a destructive interaction under the tested conditions—not a universal countdown specifying when every aluminum component will fail.
The same testing found a different outcome with other surfaces:
| Tested surface | Reported result with Conductonaut |
|---|---|
| Aluminum coldplate | Blackening, chalkiness and embrittlement |
| Nickel-plated copper heat spreader | Staining without a measurable thermal-performance loss |
| Bare copper heat spreader | Heavy staining, with performance retained in the test |
These findings apply to the tested Conductonaut installations; they are not blanket certification for every liquid-metal formulation or surface coating.
The practical lesson is straightforward: verify the actual contact surface and the compound manufacturer’s compatibility requirements before considering an application. A compatible surface is a prerequisite, not proof that the whole installation is safe.
Factory liquid metal includes engineering beyond the compound
Factory adoption is legitimate evidence that liquid metal can work in a finished product. It is not evidence that the protective design can be omitted.
ASUS describes an anti-leak fence and custom automated equipment that meters the application precisely. Its ROG engineering retrospective also explains how automated application and a surrounding barrier sponge helped bring liquid metal into its laptop designs. Those are manufacturer descriptions rather than independent reliability measurements, but they show that application and containment are part of the cooling system.
Sony offers another example. In its April 21, 2025 PS5 Pro teardown, the company explained that adopting liquid metal for the original PS5 involved safety testing and substantial insulation research. PS5 Pro retained the basic structure while adding fine grooves at the application area to stabilize cooling.
The syringe supplies the interface material. It does not supply the laptop maker’s manufacturing process or the console engineer’s insulation design.
Application instructions are product-specific
Thermal Grizzly’s Conductonaut quick-start guide calls for clean, dry, residue-free contact surfaces. It specifies a pinhead-sized amount on both the CPU and cooler, spread across both surfaces using the supplied cotton swabs. The same guide warns against aluminum contact and restricts direct-heatpipe-touch cooler use to copper bases.
Those directions explain why treating Conductonaut like an ordinary paste application is inappropriate. They are product-specific instructions, not a complete conversion procedure for an arbitrary laptop, graphics card or desktop cooler.
For factory-equipped ASUS notebooks, the manufacturer’s advice remains more conservative: do not independently dismantle the thermal module or substitute cooling materials. A retail compound’s application guide does not override device-specific service guidance.
Make it a deliberate modification, not a reflex upgrade
The editorial judgment here is not that liquid metal is useless or inherently doomed to fail. The benchmark evidence shows genuine benefits, and engineered factory installations demonstrate legitimate applications. The problem is assuming those benefits arrive without the surrounding compatibility checks, containment and servicing requirements.
Before choosing it, ask three questions: What measured problem am I solving? Does the evidence match my hardware and application location? Can I accept the consequences if the modification goes wrong?
If those answers are vague, leave the syringe alone. Lower temperatures are satisfying; turning a cooling experiment into a replacement-hardware project rather less so.