The Art of Testing Thermal Pastes: A Delicate Balance

We want to see on the CPU right so that's again the thing you can if you just do this kind of testing you can tune your paste or liquid metal or whatever that it will show higher numbers right but practically it can still be worse. They can they produce a temperature number as well but unless and I we could have them do this but unless you change the dummy heater to have say specific mosfet locations for like IO CCD whatever and also have an actual transplanted IHS onto it yep it just seems like it's it seems like you might as well do some kind of modified real world testing. Yessssss, that's what we recommend I mean in the end that would save me about seventy thousand dollars so I appreciate it yeah yeah and um I mean I would love that you hunt down the the rabbit hole of uh showing that all the conductivity numbers don't make sense yeah um I mean that's probably I'd be willing to spend that but probably a nice video I'm not sure if it's worth spending 7K 70k on it um but worth it yeah now fun probably yes yeah I mean I like these type of videos you would like to see it as well I think but uh it's maybe a bit over the top but in the end I mean in the in the end the user will only care is my CPU cooler or not right and that's also a thing um we can argue about that maybe the test methodology on like a CPU is not accurate enough so it's like introducing too many problems like mounting pressure and the temperature sensor and the CPU but I mean if you're if you're if your result is sold like close to each other that is kind of a result too I guess yeah it's a result how does it matter right like if you're the user out there and you want to figure out if this paste is 0.02 degree uh better than the other one and that's why it sounds it seems like uh from a manufacturer's perspective like thermal Grizzly the argument to make would not necessarily liquid Metals a little different yeah but for Pace would not necessarily be performance purely it would also be longevity and ease of application yeah that's why we are working on a new Pace right now um probably still one year ahead maybe more um that's trying to be as good as cry or not but just a lot better in the longevity and still being able to apply it because like over the years I just came from an Enthusiast perspective just doing ln2 all day long so I had a completely different mindset and I didn't really care what other people were saying right but now that um I mean I obviously care a lot about if the normal user can apply it or not yeah so we are trying to adapt to that and I think also for you I think the longevity and like the application is much more important than if it's one or two degrees Celsius better yeah like being being fully honest even though obviously I would love that people buy crying out extreme but being honest it doesn't make sense yeah yeah and some of that stuff too like it's okay to leave the flagship Halo for the flagship Halo user like someone on stage downstairs doing Allen too and not necessarily yourself to everyone so yeah yeah now that that makes sense yeah and uh I mean to close this conductivity topic um if you're out there just don't go by those numbers and just check reviews and if the result is so close to each other then just check what other like properties the pace has no matter which which manufacturer it is just check if it's maybe if it's cheaper maybe if it's easier to apply maybe if the comments say hey it lasts longer or whatever right I was gonna say even if it's not validated by someone so I know there's like a different of uh reviews on paste in general but even if it's not validated you can still tell a lot at least from the marketing language where if there's some base trust in the company uh like say like thermal Grizzly knocked to any company that has kind of earned like a base level yeah um you look at the marketing language and at least figure out who are they trying to sell it to so they list longevity versus they don't list longevity that does tell you something useful if no reviews exist for it so yeah yeah uh anyway temperature's fake and uh energy doesn't exist Earth is flat maybe just beat that out yeah we'll bleep that one I I bleeped uh almost your entire segment from earlier it's perfect it's just you saying yeah uh YouTube yeah I asked him if he wants to have the the green or the yellow check and yeah or actually this entire video we're censoring all of it perfect so thanks for joining thanks for having me again I hope you enjoyed it yeah go check out their Bower on YouTube and uh we'll see you all next time

The Importance of Considering Real-World Testing for Thermal Pastes

Testing thermal pastes is a delicate balance that requires careful consideration. Manufacturers often rely on test results to compare the performance of different thermal paste products, but this approach can be misleading. The key to understanding thermal paste performance lies in considering real-world testing, rather than relying solely on laboratory tests.

One of the major limitations of laboratory tests is that they are often conducted under idealized conditions. This means that the tests are designed to produce optimal results, which may not accurately reflect real-world scenarios. In contrast, real-world testing involves subjecting thermal pastes to a variety of conditions, including different temperatures, humidity levels, and system configurations.

Real-World Testing: A More Accurate Method

Real-world testing is a more accurate method for evaluating the performance of thermal pastes. This approach takes into account the complexities of real-world systems, where factors such as mounting pressure, temperature fluctuations, and system configuration can significantly impact thermal paste performance.

By conducting real-world tests, manufacturers can gain a better understanding of how different thermal paste products perform under various conditions. This allows them to make more informed decisions about which products to recommend for specific applications.

The Role of Testing in Ensuring Thermal Paste Performance

Testing plays a critical role in ensuring the performance of thermal pastes. Manufacturers must conduct thorough tests to validate their product's performance, including:

1. **Temperature testing**: Evaluating how well a thermal paste can dissipate heat under different temperature conditions.

2. **Humidity testing**: Assessing how a thermal paste performs in humid environments, which can affect its viscosity and thermal conductivity.

3. **System configuration testing**: Testing the performance of thermal pastes with different system configurations, such as motherboard layouts or CPU cooler designs.

By conducting comprehensive testing programs, manufacturers can ensure that their products meet the necessary standards for thermal paste performance.

Conclusion

In conclusion, testing thermal pastes requires a nuanced approach that takes into account real-world conditions. Manufacturers must conduct thorough tests to validate their product's performance, including temperature, humidity, and system configuration testing. By doing so, they can ensure that their products deliver optimal thermal performance in various scenarios.

"WEBVTTKind: captionsLanguage: enmaterials engineer and number one Community voted Bromance of the show dare Bauer joined us for a lengthy discussion on new products thermal engineering thermal paste misconceptions industry-wide misleading marketing and thermal conductivity we originally wanted to shoot a 15-minute video but then we kept talking a lot and the end of the shoot looked like this how how long was that um uh 50 minutes 50. yeah I didn't expect it in this video their Bowers spelled MIT OCT joins us specifically to cover misleading marketing of thermal conductivity values and thermal paste and interfaces it's a subject that Roman openly admits he's participated in with his thermal Grizzly involvement but it's something that he's decided to get away from and he'd like to see the rest of the industry do the same that starts with education and that's what he's here to provide this is one of two videos because we talked a lot enough to need to split them that we'll be uploading with their power before that this video is brought to you by Deep cool and the new zero dark series of ak-620 and ak-400 CPU coolers we previously reviewed the ak-620 and ak-400 and found them to be among a new crop of extremely competitive coolers for the price the new zero dark and zero dark plus variations move out to a blackout color design with blackout FDB fans the heatsinks otherwise have the same characteristics as those that we tested previous obviously and found to be well performed just with a fresh new look learn more at the link in the description below do we start the rant now we can rent okay yeah thermal conductivity uh what is it and why is it a fake thing that's not real I mean that's what you were saying to me earlier it's not like thermal conductivity is not real but there you know that's what you said something about like thermal conductivity is flat I don't know yeah so the thing with thermiconductivity is that you can make it whatever you want you can make it the way that it Suits You Ah that's a easy easy example is liquid metal yeah um the thermiconductivity is dependent on temperature and there is no real standard for it um a lot of the like standard testing machines are testing at 20 degrees Celsius then you can argue that's too far away from application so I tested 100 degrees Celsius and yeah which so we talk on this yes at LTX like five or six years ago I remember asking the same question uh which one produces the better number if you test it's a higher temperature okay that's what I thought but the higher temperature is more realistic yeah so not only is it more realistic but it produces a more advantageous marketing though yeah but I can also test at 500 degrees Celsius okay well why don't you do that yeah I mean and to be fully honest we did that okay yeah no not 500 another 500 but at a higher temperature yeah and um because there is no standard yeah you can I mean it's it's a curve and you don't Define at what point it's actually tested because we're also giving a temperature range I'm not sure what it is for liquid metal but it might be like up to 350 degrees Celsius or something like that so I can test at Ma I can test at -10 I can test at plus 350. right and then I can make make it look the way I want and so and here's the thing too is like we've pointed this out uh I think I think at GM we started pointing it out more after I talked to you the first time but uh hopefully most of the time comparing thermal conductivity numbers within a manufacturer should be comparable but I guess it depends on if they change how they modify that curve per product and even then I guess yeah I question maybe the value of that number yeah I just walk it back really quick but we're talking about the watts per meter Calvin number yes just for anyone who's good yes so so let's say you have a thermal paste that's showing eight what per meter Kelvin or you have this um that in theory can show you like I don't know like 1000 we obviously don't list it on there because like um there's also because people keep asking me this a lot like why is it not in a package like three or four years ago we decided that it's not a good way to advertise a product sure um I think it takes Focus away from what it's actually trying to do right yeah and um it's the same for the pace it's kind of misleading um because I think the the best way to do it is that we send it to reviewers and the reviewer is just comparing paste a versus paste b or this product and then show what the real differences well even that is difficult because we've done some limited uh thermal interface tests but it was mostly like the sheets like the icy diamond stuff uh but the one of the challenges with thermal interface is as uh obviously you know but for the audience is not only what's the performance in terms of temperature difference in the product you're testing but also you have to consider things like longevity and aging uh dry out um if it's in an extreme application how well does it withstand liquid nitrogen stuff like that and and so even reviews are very difficult to properly do it yeah and if you let's say you talk about icy Diamond I'm not sure what it is but if you look at those graphic graphic sheets yeah they often list like 600 watt per meter Kelvin and then they perform the same as a thermal paste that that's listed with eight and then you should start thinking that it's it's not a good metric doesn't mean anything um a very good example I usually uh have is if you look at a solid piece of copper and it has like 390 watt per meter Kelvin thermiconductivity right if I put 25 degrees Celsius yeah and if I if I turn this into dust and add a little bit of silicon oil it ends up at like I don't know like five Alchemist or what it turned into gold I wish I could no but if if you add a little bit of silicon oil so basically it's a paste then it turns into like an overall thermal conductivity of five to ten okay so it's a lot worse even though the base material is still the same you can work the inverse way yeah so thinking about that if you increase the particle size then you will also increase thermal conductivity as stupid as this might sound but that's the way it is but you increase particle size you increase particle size you introduce problems I would assume with mounting and application yeah where depending on how extreme you go with the particle size it seems like those would start causing contact problems yeah you increase the the layer and that's why the the thermal conductivity will be higher but you're like Delta between the let's say the CPU and the cooler will be higher right so your temperature will be worse even though your thermal conductivity will be higher right and so this way you can if you want to you can tune a paste in a way that it looks nice on a bench but it will perform bad yeah and that's why I just wanted to highlight that like like don't go after those conductivity values so just trust reviewers and uh well let me ask you this so we we haven't bought one of these yet uh I was thinking about it though so there's two types of like thermal paste testing machines one of them is much more advanced um I don't know if you know a long one the company no okay so they're a local local Taiwan manufacturer and uh they have these large thermal interface testing devices they're automated they can do all kinds of stuff but uh there's dummy heaters they apply a known amount of force they can do aging testing by running it through multiple Cycles it's very Advanced I think methodologically I would trust the machine methodologically what I don't know if I can trust and the reason I haven't bought one is how much does that matter how much does that reflect real world used or at what point because if you try to test with real components you're not going to see any difference a lot of the time that's that's that's the the thing I wanted or the thing I meant because you get values from that you can get conductivity values but they don't reflect the actual performance we want to see on the CPU right so that's again the thing you can if you just do this kind of testing you can tune your paste or liquid metal or whatever that it will show higher numbers right but practically it can still be worse and they can they produce a temperature number as well but unless and I we could have them do this but unless you change the dummy heater to have say specific mosfet locations for like IO CCD whatever and also have an actual transplanted IHS onto it yep it just seems like it's it seems like you might as well do some kind of modified real world testing yep that's what we recommend I mean in the end that would save me about seventy thousand dollars so I appreciate it yeah yeah and um I mean I would love that you hunt down the the rabbit hole of uh showing that all the conductivity numbers don't make sense yeah um I mean that's probably I'd be willing to spend that but probably a nice video I'm not sure if it's worth spending 7K 70k on it um but worth it yeah now fun probably yes yeah I mean I like these type of videos you would like to see it as well I think but uh it's maybe a bit over the top but in the end I mean in the in the end the user will only care is my CPU caller or not right and that's also a thing um we can argue about that maybe the test mythology on like a CPU is not accurate enough so it's like introducing too many problems like mounting pressure and the temperature sensor and the CPU but I mean if you're if you're if your result is sold like close to each other that is kind of a result too I guess yeah it's a result how does it matter right like if you're the user out there and you want to figure out if this paste is 0.02 degree uh better than the other one and that's why it sounds it seems like uh from a manufacturer's perspective like thermal Grizzly the argument to make would not necessarily liquid Metals a little different yeah but for Pace would not necessarily be performance purely it would also be longevity and ease of application yeah that's why we are working on a new Pace right now um probably still one year ahead maybe more um that's trying to be as good as cry or not but just a lot better in the longevity and still being able to apply it because like over the years I just came from an Enthusiast perspective just doing ln2 all day long so I had a completely different mindset and I didn't really care what other people were saying right but now that um I mean I obviously care a lot about if the normal user can apply it or not yeah so we are trying to adapt to that and I think also for you I think the longevity and like the application is much more important than if it's one or two degrees Celsius better yeah like being being fully honest even though obviously I would love that people buy crying out extreme but being honest it doesn't make sense yeah yeah and some of that stuff too like it's okay to leave the flagship Halo for the flagship Halo user like someone on stage downstairs doing Allen too and not necessarily yourself to everyone so yeah yeah now that that makes sense yeah and uh I mean to close this conductivity topic um if you're out there just don't go by those numbers and just check reviews and if the result is so close to each other then just check what other like properties the pace has no matter which which manufacturer it is just check if it's maybe if it's cheaper maybe if it's easier to apply maybe if the comments say hey it lasts longer or whatever right I was gonna say even if it's not validated by someone so I know there's like a different of uh reviews on paste in general but even if it's not validated you can still tell a lot at least from the marketing language where if there's some base trust in the company uh like say like thermal Grizzly knocked to any company that has kind of earned like a base level yeah um you look at the marketing language and at least figure out who are they trying to sell it to so they list longevity versus they don't list longevity that does tell you something useful if no reviews exist for it so yeah yeah uh anyway temperature's fake and uh energy doesn't exist Earth is flat maybe just beat that out yeah we'll bleep that one I I bleeped uh almost your entire segment from earlier it's perfect it's just you saying yeah uh YouTube yeah I asked him if he wants to have the the green or the yellow check and yeah or actually this entire video we're censoring all of it perfect so thanks for joining thanks for having me again I hope you enjoyed it yeah go check out their Bower on YouTube and uh we'll see you all next timematerials engineer and number one Community voted Bromance of the show dare Bauer joined us for a lengthy discussion on new products thermal engineering thermal paste misconceptions industry-wide misleading marketing and thermal conductivity we originally wanted to shoot a 15-minute video but then we kept talking a lot and the end of the shoot looked like this how how long was that um uh 50 minutes 50. yeah I didn't expect it in this video their Bowers spelled MIT OCT joins us specifically to cover misleading marketing of thermal conductivity values and thermal paste and interfaces it's a subject that Roman openly admits he's participated in with his thermal Grizzly involvement but it's something that he's decided to get away from and he'd like to see the rest of the industry do the same that starts with education and that's what he's here to provide this is one of two videos because we talked a lot enough to need to split them that we'll be uploading with their power before that this video is brought to you by Deep cool and the new zero dark series of ak-620 and ak-400 CPU coolers we previously reviewed the ak-620 and ak-400 and found them to be among a new crop of extremely competitive coolers for the price the new zero dark and zero dark plus variations move out to a blackout color design with blackout FDB fans the heatsinks otherwise have the same characteristics as those that we tested previous obviously and found to be well performed just with a fresh new look learn more at the link in the description below do we start the rant now we can rent okay yeah thermal conductivity uh what is it and why is it a fake thing that's not real I mean that's what you were saying to me earlier it's not like thermal conductivity is not real but there you know that's what you said something about like thermal conductivity is flat I don't know yeah so the thing with thermiconductivity is that you can make it whatever you want you can make it the way that it Suits You Ah that's a easy easy example is liquid metal yeah um the thermiconductivity is dependent on temperature and there is no real standard for it um a lot of the like standard testing machines are testing at 20 degrees Celsius then you can argue that's too far away from application so I tested 100 degrees Celsius and yeah which so we talk on this yes at LTX like five or six years ago I remember asking the same question uh which one produces the better number if you test it's a higher temperature okay that's what I thought but the higher temperature is more realistic yeah so not only is it more realistic but it produces a more advantageous marketing though yeah but I can also test at 500 degrees Celsius okay well why don't you do that yeah I mean and to be fully honest we did that okay yeah no not 500 another 500 but at a higher temperature yeah and um because there is no standard yeah you can I mean it's it's a curve and you don't Define at what point it's actually tested because we're also giving a temperature range I'm not sure what it is for liquid metal but it might be like up to 350 degrees Celsius or something like that so I can test at Ma I can test at -10 I can test at plus 350. right and then I can make make it look the way I want and so and here's the thing too is like we've pointed this out uh I think I think at GM we started pointing it out more after I talked to you the first time but uh hopefully most of the time comparing thermal conductivity numbers within a manufacturer should be comparable but I guess it depends on if they change how they modify that curve per product and even then I guess yeah I question maybe the value of that number yeah I just walk it back really quick but we're talking about the watts per meter Calvin number yes just for anyone who's good yes so so let's say you have a thermal paste that's showing eight what per meter Kelvin or you have this um that in theory can show you like I don't know like 1000 we obviously don't list it on there because like um there's also because people keep asking me this a lot like why is it not in a package like three or four years ago we decided that it's not a good way to advertise a product sure um I think it takes Focus away from what it's actually trying to do right yeah and um it's the same for the pace it's kind of misleading um because I think the the best way to do it is that we send it to reviewers and the reviewer is just comparing paste a versus paste b or this product and then show what the real differences well even that is difficult because we've done some limited uh thermal interface tests but it was mostly like the sheets like the icy diamond stuff uh but the one of the challenges with thermal interface is as uh obviously you know but for the audience is not only what's the performance in terms of temperature difference in the product you're testing but also you have to consider things like longevity and aging uh dry out um if it's in an extreme application how well does it withstand liquid nitrogen stuff like that and and so even reviews are very difficult to properly do it yeah and if you let's say you talk about icy Diamond I'm not sure what it is but if you look at those graphic graphic sheets yeah they often list like 600 watt per meter Kelvin and then they perform the same as a thermal paste that that's listed with eight and then you should start thinking that it's it's not a good metric doesn't mean anything um a very good example I usually uh have is if you look at a solid piece of copper and it has like 390 watt per meter Kelvin thermiconductivity right if I put 25 degrees Celsius yeah and if I if I turn this into dust and add a little bit of silicon oil it ends up at like I don't know like five Alchemist or what it turned into gold I wish I could no but if if you add a little bit of silicon oil so basically it's a paste then it turns into like an overall thermal conductivity of five to ten okay so it's a lot worse even though the base material is still the same you can work the inverse way yeah so thinking about that if you increase the particle size then you will also increase thermal conductivity as stupid as this might sound but that's the way it is but you increase particle size you increase particle size you introduce problems I would assume with mounting and application yeah where depending on how extreme you go with the particle size it seems like those would start causing contact problems yeah you increase the the layer and that's why the the thermal conductivity will be higher but you're like Delta between the let's say the CPU and the cooler will be higher right so your temperature will be worse even though your thermal conductivity will be higher right and so this way you can if you want to you can tune a paste in a way that it looks nice on a bench but it will perform bad yeah and that's why I just wanted to highlight that like like don't go after those conductivity values so just trust reviewers and uh well let me ask you this so we we haven't bought one of these yet uh I was thinking about it though so there's two types of like thermal paste testing machines one of them is much more advanced um I don't know if you know a long one the company no okay so they're a local local Taiwan manufacturer and uh they have these large thermal interface testing devices they're automated they can do all kinds of stuff but uh there's dummy heaters they apply a known amount of force they can do aging testing by running it through multiple Cycles it's very Advanced I think methodologically I would trust the machine methodologically what I don't know if I can trust and the reason I haven't bought one is how much does that matter how much does that reflect real world used or at what point because if you try to test with real components you're not going to see any difference a lot of the time that's that's that's the the thing I wanted or the thing I meant because you get values from that you can get conductivity values but they don't reflect the actual performance we want to see on the CPU right so that's again the thing you can if you just do this kind of testing you can tune your paste or liquid metal or whatever that it will show higher numbers right but practically it can still be worse and they can they produce a temperature number as well but unless and I we could have them do this but unless you change the dummy heater to have say specific mosfet locations for like IO CCD whatever and also have an actual transplanted IHS onto it yep it just seems like it's it seems like you might as well do some kind of modified real world testing yep that's what we recommend I mean in the end that would save me about seventy thousand dollars so I appreciate it yeah yeah and um I mean I would love that you hunt down the the rabbit hole of uh showing that all the conductivity numbers don't make sense yeah um I mean that's probably I'd be willing to spend that but probably a nice video I'm not sure if it's worth spending 7K 70k on it um but worth it yeah now fun probably yes yeah I mean I like these type of videos you would like to see it as well I think but uh it's maybe a bit over the top but in the end I mean in the in the end the user will only care is my CPU caller or not right and that's also a thing um we can argue about that maybe the test mythology on like a CPU is not accurate enough so it's like introducing too many problems like mounting pressure and the temperature sensor and the CPU but I mean if you're if you're if your result is sold like close to each other that is kind of a result too I guess yeah it's a result how does it matter right like if you're the user out there and you want to figure out if this paste is 0.02 degree uh better than the other one and that's why it sounds it seems like uh from a manufacturer's perspective like thermal Grizzly the argument to make would not necessarily liquid Metals a little different yeah but for Pace would not necessarily be performance purely it would also be longevity and ease of application yeah that's why we are working on a new Pace right now um probably still one year ahead maybe more um that's trying to be as good as cry or not but just a lot better in the longevity and still being able to apply it because like over the years I just came from an Enthusiast perspective just doing ln2 all day long so I had a completely different mindset and I didn't really care what other people were saying right but now that um I mean I obviously care a lot about if the normal user can apply it or not yeah so we are trying to adapt to that and I think also for you I think the longevity and like the application is much more important than if it's one or two degrees Celsius better yeah like being being fully honest even though obviously I would love that people buy crying out extreme but being honest it doesn't make sense yeah yeah and some of that stuff too like it's okay to leave the flagship Halo for the flagship Halo user like someone on stage downstairs doing Allen too and not necessarily yourself to everyone so yeah yeah now that that makes sense yeah and uh I mean to close this conductivity topic um if you're out there just don't go by those numbers and just check reviews and if the result is so close to each other then just check what other like properties the pace has no matter which which manufacturer it is just check if it's maybe if it's cheaper maybe if it's easier to apply maybe if the comments say hey it lasts longer or whatever right I was gonna say even if it's not validated by someone so I know there's like a different of uh reviews on paste in general but even if it's not validated you can still tell a lot at least from the marketing language where if there's some base trust in the company uh like say like thermal Grizzly knocked to any company that has kind of earned like a base level yeah um you look at the marketing language and at least figure out who are they trying to sell it to so they list longevity versus they don't list longevity that does tell you something useful if no reviews exist for it so yeah yeah uh anyway temperature's fake and uh energy doesn't exist Earth is flat maybe just beat that out yeah we'll bleep that one I I bleeped uh almost your entire segment from earlier it's perfect it's just you saying yeah uh YouTube yeah I asked him if he wants to have the the green or the yellow check and yeah or actually this entire video we're censoring all of it perfect so thanks for joining thanks for having me again I hope you enjoyed it yeah go check out their Bower on YouTube and uh we'll see you all next time\n"