Carbon Nanotube CPU Cooling
Recorded: May 31, 2026, 3 a.m.
| Original | Summarized |
Carbon Nanotube CPU Cooling with Carbice | LTT LabsAll ProductsCategoriesArticlesVSVersusModeComparison Bin←BACK TO ARTICLESTable of ContentsCarbon NanotubesCarbiceStructureInterfaceAdaptabilityApplicationsManufacturing and SpaceConsumersGraphite Pads 2?TestingComparisons, Application, and ActivationCarbice Ice PadPTM7950Noctua NT-H2Test System and ProcedureResultsTIM ComparisonResultsOutlierRemoval DamageReduce, Reuse, RecycleShould I Reuse My Ice Pad?More on ActivationWho is this for? Does it solve their problem?Fun FAQSWill I damage it if I apply too much pressure with my cooler?Are the carbon nanotubes easy to damage?Does it bend, and can you use it if it's bent?Are there any differences between Ice Pad and Space Pad?(not actually frequently asked, but an excuse for a fun fact)Carbon Nanotube CPU Cooling with CarbiceLucas N.·Tested by Jon D., and Sean H.·Carbon Nanotube CPU Cooling with CarbiceCarbon Nanotube CPU Cooling with CarbiceLucas N.·Tested by Jon D., and Sean H.·Table of ContentsExplore similarExplainerCPUDoes using a cool material technology actually help keep cool your CPU? Carbice sure hopes it does, but they didn't rely on those hopes when designing their Space Pad for operation in orbit and the Ice Pad for use in consumer electronics.Carbice is a company founded in 2011 that manufactures aluminum sheets populated with forests of Carbon NanoTubes(CNT) for use as a Thermal Interface Material(TIM), the scientific term for what thermal paste does. Carbice’s pads can be used to thermally bond surfaces(small or large) and facilitate heat transfer while avoiding the tedious application and cleanup of a thermal compound or liquid metal.Carbice visited our office to drop off some samples and show us the product. Carbice wanted to show us this cool technology and we wanted to learn about it, but this testing and article is not sponsored, we received nothing but the samples and very interesting answers to our questions. Let’s get into the theory of what they showed us, and some of our own testing!Carbice Visit LTT LabsCarbon NanotubesCarbon nanotubes are a sensational material/nanostructure that have been developed through the second half of the 20th century and into the 21st century. They've been explored for use in many fields due to their great potential for thermal conductivity, mechanical strength, and various electrical properties. There has been a lot of interest and you’ve likely seen headlines about how much stronger they are than steel, but being an exciting new technology isn't enough, consistent manufacturing and practical benefits are required for viable implementations.Rotating Single-Wall Carbon NanotubeSource: https://en.wikipedia.org/wiki/Carbon_nanotubeAs carbon can be formed into diamond or graphite sheets, carbon can also be formed into single-walled nanoscale tubes.(illustrated above) These tubes are at the scale of single-nanometers, and can also be built up to have multiple layers of walls(like in Carbice pads). The individual CNT are fragile on their own, but they can also be grown in a forest structure, providing mutual support like bristles of a brush.Like all topics in the field of nano-technology, there are an incredible number of in-depth papers that can be found about nanotubes. These go multiple PhDs deep into the math, physics, and chemistry of carbon nanotube fabrication and characterization, which is often too far if you don't already study nanotechnology. If you're looking for more information, we've found Wikipedia and the paper Carbon nanotubes: properties, synthesis, purification, and medical applications by Eatemadi, et al. (2014) to be good introductions to the hard science.The carbon nanotube property most relevant to Carbice is thermal conductivity. Similar to electrical conductivity, which describes how easily electricity can pass through a material, thermal conductivity describes how easily heat energy passes through a material. Electrical and thermal resistivity are just the inverse of conductivity, representing how difficult it is for electricity or heat to be conducted. While the thermal resistivity(property of a material) of CNTs themselves is incredibly low, the thermal resistance(property of a material implementation) can be drastically higher due to the low contact area of the CNT tips. This requires more engineering to address, but other features of CNTs, like their flexibility, have a part in the solution.CarbiceCarbice 90 μm Ice PadStructureCarbon nanotubes cannot just be spread on your CPU like a thermal paste, they require structure and alignment for effective thermal transfer. Carbice's solution is to use them in conjunction with an aluminum substrate and polymer coating. Carbice founder Baratunde A. Cola published a paper in 2007 forming the basis of Carbice's CNT manufacturing process: CNT synthesis on both sides of an aluminum foil.Using a centre of 50 μm thick aluminum substrate, they deposit CNT forests on both sides using a Chemical Vapour Deposition(CVD) process. Then a polymer coating is applied to both surfaces, later impregnating the CNT forests during activation. This creates a symmetrical stack-up as illustrated below.Carbice Pad StructureSource: https://carbice.com/technology/how-it-worksThis symmetrical structure has the carbon nanotubes anisotropically transfer the heat from the hot interface to the aluminum substrate where there can be some lateral heat transfer. The heat is then transferred through the other layer of nanotubes to the cold interface. Carbice has said that they create CNT forests in thicknesses of 5 μm to hundreds of um, though for CPU applications they typically range from 20-50 μm. This results in pads that are 90-150 μm thick(aluminum substrate and two layers of CNTs).Large CPUs and other applications with greater curvature can benefit from thicker pads to better conform to surfaces and account for gaps. These applications can be addressed with thicker CNT forests, but they can also be addressed by stacking two thinner pads together. The best solution is determined by Carbice based on simulations and in-application testing, considering compressibility, contact, and overall performance.Conformity of Carbice PadsSource: https://carbice.com/technology/assemblyInterfaceOne downside of CNT forests is the thermal contact resistance between the tips of the CNTs and the surfaces of the hot and cold plates(the surfaces they're transferring heat from/to). The tips of the CNT can 'bottleneck' the transfer of heat to the Carbice pads due to the low contact surface area.This is partially addressed by the polymer coating that is applied to both surfaces of the Carbice pad, requiring an activation procedure, but vastly reducing the thermal resistance. It acts as a sort of thermal interface material for the thermal interface material, 'wetting' the CNT forest and other surfaces to increase the contact area.AdaptabilityTo further address the problem of low surface area and high thermal resistance of CNT tips, the CNTs will buckle and move, bending to the contour of the adjacent surface and taking up the slack of surface roughness or concavity. With time and repeated heat cycles, the CNT will move and settle into place, increasing surface area and reducing thermal resistance. Carbice has conducted testing and found their solutions to outperform thermal paste in the long term as this settling occurs and there is no pump out or dry out. Carbice CNT Spring ActionSource: https://carbice.com/technology/how-it-worksApplicationsManufacturing and SpaceCarbice entered the market as a thermal solution for the manufacturing and space industries, and one of the greatest benefits of Carbice's products is how simple they are to apply, remove, and replace. Carbice is much simpler to integrate into an automated assembly line, requiring only peeling of a protective film and then application onto the desired surface. The polymer coating can even be treated with an adhesive for ease of use.If the Carbice pad has to be removed for repair or rework, then it can easily be peeled off. In our experience it usually left some CNTs behind on the surface, but these can be scrubbed away with isopropyl alcohol and light pressure. While the aluminum and carbon pads are (very) electrically conductive, they don't spread or have corrosive effects like liquid metal.This is great for space industry applications where a satellite may be assembled, tested, and taken apart again many times before the final product. Additionally, Carbice has told us about their state-of-the-art simulation tools which allow them to evaluate solutions before they're implemented. These can drastically reduce the amount of expensive testing and iteration required.ConsumersCarbice is now bringing their CNT technology to desktop computers, first as a collaboration with CyberPowerPC in August 2025. The Ice Pad is available as a $10 configuration option on customizable configurations through CyberPowerPC, but it is yet to be available to consumers as a standalone product. CyberPowerPC also has a FAQ page for the Carbice Ice Pad.CyberPowerPC ConfiguratorSource: https://www.cyberpowerpc.com/We haven't been able to find many user stories or discussion of the Ice Pad in CyberPowerPC systems on the internet. This could mean that few people are choosing the option, or it is performing well in these systems and there have been little to no issues. It seems like an intriguing and low cost build option so we imagine it’s the latter. If you, or someone you know, has chosen this option then please reach out!Graphite Pads 2?For anyone passionate about CPU cooling solutions, this explanation will have inspired some questions about similarities and differences to graphite pads that were popular years ago. These Carbice pads do have similarities, though Carbice’s claim is that the ability for the CNT forests to conform and adapt to the undulations of the surface will allow greater contact, thus greater performance. This should position it against performance TIM solutions like high end thermal pastes and phase change materials, so that is how we tested it.The Ice Pads are receiving significant interest from system integrators like graphite pads once did, and so far they appear to be holding up to their scrutiny; we will have to see where this goes.TestingComparisons, Application, and ActivationAs comparisons to the Carbice Ice Pad, we also tested PTM7950 thermal pads and Noctua NT-H2 thermal compound. The PTM7950 represents the 'easy to apply' performance alternative while the NT-H2 is a premium thermal compound, a format familiar to PC building enthusiasts. They all had slightly different preparation and testing procedures.Noctua NT-H2, Carbice Ice Pad, and PTM7950Carbice Ice PadWhen conducting performance testing, Carbice Ice Pad installation requires an ‘activation’ procedure where both sides of the thermal pad must reach 33-37°C for a short period of time(at least a few minutes) for the polymer coating to activate and spread, enabling the expected performance. This speeds up the process of the polymer coating diffusing into the CNT forest, and the CNT conforming to the other surfaces.This dedicated activation procedure shouldn’t be required for use by regular users of the Ice Pads, and regular temperature cycling will cause this activation to happen with use. Carbice reports this as part of the reason they see Ice Pad performance improve over time.When Carbice visited, we conducted activation procedures consisting of at least 15 minutes of OverClock Checking Tool Linpack 2012 16GB(OCCT) until temperatures plateaued; we continued using this procedure for the remainder of our standard testing. An alternate activation procedure was attempted in later testing.1 of 2Carbice Ice Pad ApplicationPTM7950Like the Carbice Ice Pad, PTM7950 requires an activation heat cycle so that it can flow and make optimal contact with both surfaces. An OCCT run of 15 minutes was performed after each application of PTM7950 to ensure that it had received this heat cycle.1 of 2PTM7950 ApplicationNoctua NT-H2For the NT-H2 tests, we applied a line of thermal paste according to our standard procedure, and for the sake of fairness we also conducted a 15 minute OCCT run with each application.1 of 2Noctua NT-H2 ApplicationTest System and ProcedureSince some TIMs perform better on air cooling compared to liquid cooling setups, we set up an AMD Ryzen 9 9950X3D AM5 test system with ASUS ROG Strix X670E-F GAMING WIFI motherboard twice, once with a Noctua NH-D15 G2, and again with an Arctic Liquid Freezer III 360. To check repeatability, we applied and tested each TIM twice on each cooler, a total of four applications when considering the two cooling solutions.For both the NH-D15 and Freezer III 360, we set all fans and the water pump to a fixed 100%, conducting all tests in our AES SCH-512-4 environmental chamber set to 20°C. The system is only being used to generate heat for this test, but all chipset and GPU drivers were updated, EXPO/XMP enabled, ReBAR enabled, the iGPU was disabled, and we used our standard Windows 11 base install.We used HWInfo to collect system information and temperatures, recording the "CPU (Tctl/Tdie)" field which records the highest temperature measured on the die. We also monitored CPU package power and CPU clocks to confirm they were the same between tests.Test System in Environmental ChamberResultsTIM ComparisonPerforming short term testing on our AM5 AMD Ryzen 9 9950X3D system, we achieved the results in the table below. The two separate test runs are listed along with an average of the two runs. Cooler Computer Load Average (Run 1 | Run 2) [°C] NT-H2 Freezer III 360 NH-D15 G2 Freezer III 360 NH-D15 G2 Cooler Computer Load Average [°C] Difference to NT-H2 [°C] NT-H2 Freezer III 360 NH-D15 G2 Freezer III 360 NH-D15 G2 ResultsAs with most thermal testing, more efficient and effective TIMs should result in lower measured temperatures. Thermal testing of this scale is subject to variance from the applied load, TIM application, cooling method, and when considering the Carbice Ice Pad, if the material was sufficiently ‘activated’. Testing with different coolers and setups can lead to different results; these results are an indication of performance, but they are non-exhaustive. Our testing above shows that while NT-H2 and PTM7950 perform similarly, the Carbice Ice Pad consistently results in 4-8°C higher CPU temperatures. The typical gaming scenario of F1 24 resulted in a 4-7°C difference while the high load OCCT runs appeared to result in a larger 6-8°C swing(excluding the outlier).This difference shows that the Carbice Ice Pad doesn't have equivalent performance in our testing. However, all three TIMs offer sufficient cooling to keep the processor from reaching its maximum temperature where performance might throttle, and the other benefits of the Ice Pad may be worth this difference. The most important goal of TIMs is to prevent thermal throttling, and this is also the focus of CyberPowerPC's Ice Pad advertising.CyberPowerPC Carbice PageSource: https://www.cyberpowerpc.com/page/carbice/Carbice doesn’t claim that CNT TIMs have the best "T0"(first test) performance, though their own testing has shown more of a 1-5°C temperature difference on the first tests. Their focus is that performance can increase over time as the pad undergoes further heat cycles and the CNT forest conforms to the hot and cold surfaces. This product isn't designed for impressive initial results, but for long term stability as other TIMs experience pump-out and dry-out.We performed some basic heat cycle testing with 4 minutes of OCCT followed by 2 minutes of rest, for ten cycles. Seeing consistent temperatures within 1-2°C across the 10 cycles, but it is insufficient to draw any conclusions. Carbice has conducted their own long term testing but of course consumers will want to see independent long term testing. If there is interest, then it may warrant us setting something up.OutlierThe most notable results are the two OCCT tests of the Carbice Ice Pad with the Arctic Liquid Freezer III 360, where there was a test with a 7°C higher average temperature.This outlier result was collected with the same Carbice sample that we used for the reuse testing(discussed later), and it did show some improvement with reuse. This could be the result of a more gradual activation and improvement over time(with heat cycles) that Carbice advertises.We've shared the outlier result with Carbice and they're not sure exactly what caused this but they've suggested that our activation procedure could have been insufficient for the polymer layer to properly activate.Carbice suggested that we unplug the pump of the AIO to allow the CPU to reach higher temperatures and increase the chance of proper activation. We conducted another set of tests using this activation procedure and measured the Ice Pad to be 4°C above average temperatures achieved by the NT-H2 and PTM7950. This is in the range that we expected for the "T0" testing of the Ice Pad.Removal DamageAs with most TIMs, Carbice pads create some adhesion between the two cooling surfaces, and it can require some force or tilting to 'break the seal'. This is likely due to the applied polymer coating or adhesive that adheres further with the activation. It can lead to the Ice Pad sticking to either side of the interface and deforming upon removal. In some lucky cases, you can get an impression of the IHS you're cooling, but others result in carbon nanotube loss.1 of 2Carbice Removal DamageThe amount and severity of the CNT forest loss is dependent on many factors like the heat pattern, temperatures reached, and if the cooler is removed while the system is still warm. In general, Carbice doesn't advise reuse of installed Ice Pads. This is an area where graphite pads can take a win, offering reusability(though extent can vary) in non-performance critical applications.But, what if we just tried reusing an Ice Pad anyway?Reduce, Reuse, RecycleAfter the second test run of the Carbice Ice Pad and Arctic Liquid Freezer III 360, we had to try removing the cooler and then putting it back on again to run another test. After that test, surely if some user is reusing the sheet, they'll forget the original orientation, they might rotate it to some random direction. Then after the single rotation, why not, why don't we try to rotate the sheet a few more times and try flipping it to the other side?With the above test “plan", we performed the activation procedure again after each reinstallation, followed by an OCCT run to collect average temperatures. The test runs and results are displayed in the table below. Application Application Note Average Temperature(OCCT) [°C] Difference to Initial Application(OCCT) [°C] 1 Initial 84 - 2 Clean and Reinstall 81 -3 3 Rotate 90° 78 -6 4 Rotate 90° 83 -1 5 Rotate 90° 83 -1 6 Rotate 90° and Flip 81 -3 We will stress again that Carbice does not make any reusability claims for the Ice Pad. However, these preliminary results are relatively promising and show a slight improvement, or at least not a huge degradation, in thermal performance.We also took photographs of the sheet after each removal.1 of 9First ApplicationShould I Reuse My Ice Pad?The difficulty with recommending reuse of the Ice Pads is that success will be greatly dependent on the state of the sheet, and which sections have worn off. Carbice explained that in some cases there can be a loss of up to 10% of the CNT without a major performance impact, but if the CNT loss is over a hot spot then it can have an outsized effect on component temperatures.Within our own testing the Ice Pad which we applied and re-applied numerous times had very little degradation, but other sheets had 15-20% of the CNT forest removed after the first application. The latter case likely wouldn’t perform well upon re-test, especially if that is an area of concentrated heat.The conclusion that we'd draw from this testing is that you might be able to reuse the pad if you're doing some quick testing or are intending to watch performance closely, but otherwise Carbice recommends against reuse. While proper application of the Carbice Ice Pad can apparently result in great long term performance, we can't make any definitive statements or conclusions about how the reuse of pads will perform as time goes on.More on ActivationAs covered a couple times in this article, the Carbice Ice Pad will sufficiently transfer heat immediately after application, but for maximum performance, they must undergo some number of high temperature heat cycles. This allows the polymer layer to wick into the carbon nanotube forest, and for the carbon nanotubes to move, conforming to the surface contours.The recommendation from Carbice is that while the deliberate activation procedure is required for performance testing, where it will be installed and immediately tested, the normal consumer who receives a Carbice Ice Pad in a prebuilt doesn't need to conduct this activation procedure. The Ice Pad will be activated when the system is built by the system integrator, or as you’re setting up your computer and installing games.Who is this for? Does it solve their problem?As with most things, this is ultimately for the prospective business or consumer to decide, so hopefully the results and exploration above were helpful, and hopefully we continue to see more test results and consumer experiences with the Ice Pad. It has shown enough performance in our testing to prevent thermal throttling on some of the most powerful CPUs on the market.We have to acknowledge that Carbice could be using the “better over time” claim to deflect subpar short-term performance testing, so time will have to tell as it gets into more people's systems and there are long term test results. Carbice has been shipping Carbice Pads and seeing success with commercial applications since 2018, so there are some carbon nanotube TIM solutions out there that have been running in production (presumably without issue) for many years. There should be more information coming out about this product and it's optimal use-cases in the near future. From the test results we've seen, the Ice Pad does provide a long term performance and consistency benefit where other TIM performance can degrade.Carbice's partnership with CyberPowerPC appears to be going well since the August 2025 release, and from the rumblings we've heard, Carbice is continuing to invest in the consumer product market through system integrators and other collaborations. Throughout our work and communication with them they've been eager for our feedback and thoughts; they appear to be interested in developing a solution that works for casuals and enthusiasts alike. If you have any questions, or currently have a CyberPowerPC with a Carbice Ice Pad in it then please let us know on the LinusTechTips Forum, Reddit, or Hacker News!Fun FAQSTo finish this off, we've assembled our own 'fun FAQS'.1 of 2Carbice Ice Pad after removal. Carbice makes no claims of reusability.Will I damage it if I apply too much pressure with my cooler?No, the CNTs and aluminum substrate are able to survive and conduct heat at up to 50,000 psi.Are the carbon nanotubes easy to damage?It is resilient to most careful handling. You can touch/grab the top/bottom of the Ice Pad or CNT with the only issue being the transmission of your finger oils. The CNT won't fall off from a touch, but can be scraped off with a fingernail.Does it bend, and can you use it if it's bent?The stiffness of the 90 μm aluminum substrate feels like aluminum foil, while the 180 μm pads we had were like folding aluminum foil over a couple times. It will bend if you apply enough force, but otherwise you don't have to be overly careful.As a general rule, if it is deformed then it shouldn't be used, but it depends on the severity. If it just has a gentle curve to it then it is likely fine, but sharp creases may cause issues. Cooler mounting forces are pretty high, so they're able to flatten it out pretty well.Are there any differences between Ice Pad and Space Pad?(not actually frequently asked, but an excuse for a fun fact)Ice Pad has a polymer coating as previously covered in the article, but space applications are subject to different requirements, specifically outgassing limits. Products used in space are subject to vacuums that can draw gasses out of materials, which then condense on other sensitive components. These films interfere with operation, measurements, or even scientific photography.While the Ice Pad uses the highest performance polymer coating, Space Pad has to make a tradeoff to meet the outgassing requirements and therefore has slightly lower thermal performance. Related Articles CPUInvestigativeExplainerSpooky Scary Skeleton... Insides of a CPU 💀By Onie T.By Onie T.Would you dare pry open your own skull to make your brain run more efficiently? That’s essentially what delidding a CPU is like - an eerie operation that might send chills down your spine! 🎃CPU, Investigative, ExplainerCPULTT CompanionExplainerWhat is the right wattage for the Ryzen 7 9700X?By Onie T., Jon D.By Onie T., Jon D.When we subjected the Ryzen 5 9600X and Ryzen 7 9700X to our testing process for the LTT video for this launch, Labs identified a noteworthy observation: the Ryzen 7 9700X appears to be better suited for a higher TDP classification. CPU, LTT Companion, ExplainerCPUGamingIntel Binary Optimization: Worth it for the few?By Nikolas H., Steven N.By Nikolas H., Steven N.Intel provides a suite of software optimizations in the form of Intel Extreme Tuning Utility (XTU), Intel Application Optimizer (APO), and the new Intel Binary Optimization Tool (IBOT). Here we take a look at which ones are worth taking the time to enable.CPU, GamingExplainerDisplayApple CMF 2026 and Studio Display XDR Test ResultsBy Jon D., Lucas N.By Jon D., Lucas N.Apple has recently released their new Studio Display and Studio Display XDR, two 27-inch 5K Retina displays available with their signature $400 stands. We'll cover some of the measurements below, but a notable release alongside these displays is the new "Apple CMF 2026".Explainer, DisplayExplainerPSUATX Power Supply TimingsBy Lucas N.By Lucas N.A look at the requirements for power supplies that comply to the ATX specification along with a test and visualization we've developed.Explainer, PSUExplainerInvestigativeLTT CompanionHow Close is Too Close? Applying Fundamental Fluid Dynamics Research Methods to PC CoolingBy Louis M. Edelman, Ph.D.By Louis M. Edelman, Ph.D.In April 2025, LinusTechTips visited NASA's Langley Research Center at the invitation of the Flow Physics and Control Branch to seek an answer to the question: how far does a case fan have to be from a flat surface?Explainer, Investigative, LTT CompanionExplore similarExplainerCPUSubscribe with your favorite feed reader:RSSAtomJSONSubscribe to the Labs Newsletter(coming soon)Back to topEthics StatementCareersHelpAbout UsFeedsThis website uses affiliate links. Purchases made through these links may provide compensation to Linus Media Group.This site works best in Firefox© 2026 LTT Labs. All rights reserved.TermsPrivacyCookies |
Carbon Nanotube CPU Cooling with Carbice introduces a novel thermal interface material (TIM) solution based on carbon nanotubes (CNT) designed to facilitate efficient heat transfer by eliminating the need for traditional thermal compounds or liquid metal. Carbice manufactures aluminum sheets infused with CNT forests, which serve to thermally bond surfaces and enhance heat transfer across the interface. The fundamental principle relies on the unique properties of carbon nanotubes, which possess high potential for thermal conductivity, mechanical strength, and electrical characteristics. While individual CNTs are fragile, they can be grown into forest structures that provide mutual support, and their thermal performance is intrinsically linked to the contact area between the nanotubes and the surfaces they interface with. Carbice’s approach involves creating a symmetrical structure where CNT forests are deposited on both sides of an aluminum substrate using Chemical Vapour Deposition (CVD) and subsequently coated with a polymer layer. This structure allows the CNTs to anisotropically transfer heat from the hot interface to the aluminum substrate, facilitating lateral heat transfer, and then transferring the heat across the other layer to the cold interface. The structure is often achieved with CNT forests ranging from several micrometers to hundreds of micrometers in thickness, typically resulting in pads that are ninety to one hundred fifty micrometers thick (including the aluminum substrate). A significant challenge addressed by Carbice is the low thermal contact resistance presented by CNT tips due to limited contact area. This issue is mitigated by the application of the polymer coating, which acts as a thermal interface material for the TIM itself, effectively 'wetting' the CNT forest and increasing the overall contact area. Furthermore, the flexibility of the CNTs allows them to buckle and move, conforming to surface contours and accounting for surface roughness or concavity. Through repeated heat cycles, the nanotubes are theorized to settle into place, increasing the effective surface area and consequently reducing thermal resistance, potentially outperforming conventional thermal pastes over the long term because they avoid the issues of pump-out or drying out experienced with liquid metals. The performance of Carbice pads is heavily dependent on an activation procedure. The polymer coating requires heat cycling, specifically reaching temperatures between thirty-three and thirty-seven degrees Celsius for a period, to activate and allow the polymer to diffuse into the CNT forest, enabling the nanotubes to conform to the surfaces. Although this activation is necessary for initial maximum performance testing, Carbice suggests that regular thermal cycling during use can lead to performance improvement over time as the CNTs settle. The applicability of this technology spans various sectors, including manufacturing and space applications, where the ease of assembly and removal of the polymer-coated sheets into automated lines is advantageous, and the non-corrosive nature of the material is beneficial for environments like space exploration. For consumer applications, Carbice is bringing the technology to desktop computers. Comparisons with established thermal solutions, such as PTM7950 thermal pads and Noctua NT-H2 thermal compound, were conducted to assess performance. When testing with system components under various loads, such as the OCCT benchmark, the Carbice Ice Pad generally resulted in CPU temperatures that were four to eight degrees Celsius higher than those measured with established TIMs. However, the focus of testing and Carbice's claim is on long-term stability and preventing thermal throttling, rather than achieving the absolute lowest initial temperature reading. In evaluating the physical integrity, the process of removal can create adhesion between the pad and the cooling surfaces, and the loss of CNTs is dependent on heat patterns and removal conditions. While Carbice does not claim specific reusability, preliminary tests involving repeated application and rearrangement of the Ice Pad showed minor degradation, suggesting that while caution is advised against reuse, slight performance retention may be observed if managed carefully. The research points toward the potential for performance gains over time due to the material’s ability to adapt to thermal stresses, and continued investigation is required to establish definitive long-term performance benchmarks independent of initial testing procedures. |