**The Science Behind Cooling Systems: A Deep Dive**

Dropping data into the spreadsheet and producing delta T over ambient values is a crucial step in understanding cooling systems. In instances where two devices are precisely controlled and can be directly compared with their base results, delta T over ambient is used as a simple subtraction of the ambient temperature from the temperature measured at the die level or component level. This scaling mostly linearly for the first at least couple degrees, maybe five to ten or so, but after that it stops scaling linearly.

For CPU core voltage, must be fixed, as this is the biggest factor when it comes to invalid results. All minor voltages should also be fixed, and fan position and cooler control are essential. Ensuring that fans are in the exact same vertical position for all tests is crucial, as moving them even one inch in either direction can influence the results by a couple of degrees depending on where it's positioned. To achieve this, an outline of the cliffs on the cooler was drawn to place them in the same spot.

To control the thermal compound used, a synthetic compound was manually spread to ensure consistency. Although better methods are available, such as blender 2.79, this method ensures accuracy and reliability. The GM logo is used for the CPU load, which monitors power consumption every millisecond of the test to determine if it's the same for all tests.

The render is held for approximately 23 minutes, sufficient to reach steady state after just the first couple of minutes. To ensure accuracy, averaging from steady state and using multiple test passes are crucial. The temperature reported as TDI (Thermal Diode Information) provides valuable data on CPU performance. However, conclusions drawn from earlier claims about the NHED 14s outperforming the NHD 15 due to heat pipes alone cannot be replicated.

These claims were based on an open-air bench setup and might have been influenced by positioning, case obstructions, or other factors that could affect cooling performance. The experiment was designed to test various cooling systems under controlled conditions, but it's essential to note that not all scenarios may yield the same results. While this experiment did not validate claims about heat sources and alignment affecting overall cooling performance, further testing might be necessary to confirm these findings.

In general, when building a PC or working with cooling systems, there are more critical factors to consider than just the die-to-die temperature difference. Ensuring that components fit within the case and air intake is properly positioned can make a significant difference in overall system performance. While precision temperature measurements are essential, they should not be the sole focus when building a PC or evaluating cooling systems.

In conclusion, the science behind cooling systems is complex, and various factors can affect performance. By understanding the intricacies of temperature measurement and control, engineers and enthusiasts can design more efficient cooling systems that provide optimal performance for their specific needs. Whether you're a seasoned gamer or an enthusiast looking to improve your PC build, considering these principles will help you make informed decisions about cooling systems.

**Additional Resources**

For those interested in learning more about PC building and cooling systems, the following resources are available:

* **Storlock Gamers XS Net**: A community-driven forum for enthusiasts to discuss PC building, gaming, and related topics.

* **Patreon.com/scishow**: A platform offering exclusive content, including articles, videos, and podcasts, on various science-related topics.

* **Blender 2.79**: A free and open-source 3D modeling software used in this experiment for precise temperature control.

**Note:** The article is based on the provided transcription and may not be suitable for general consumption without proper understanding of the technical terms and concepts discussed.

"WEBVTTKind: captionsLanguage: enbakwin rise in 3000 launched there was reasonable speculation founded in basic physics that the asymmetrical dye arrangement of the CPUs with fewer chip 'lets could have implications for cooler performance the idea was that at the root of it a cooler whose heat pipes aligned to fully contact above the dye would perform better as opposed to one with two heat pipes sharing vertical contact with the dye we still see a lot of online commentary about this and some threads about which orientation of a cooler is best so we thought we'd bust a few of the myths that popped up but also do some testing on the base idea before that this video is brought to you by the msi gtx 1660 super gaming x the gtx 1660 super is a strong mid-range performer that's capable of 1440p gaming without issue and MSI's gaming experience uses a dual fan cooler with a more traditional two slot design for versatility it also has our RGB LEDs of course learn more at the link below this is pretty old news by now with much of the original discussion starting about two months ago but it's still alive and noctua recently revives this issue at the end of october knocked over knocked to a server at the end of october by stating that it believed there to be no meaningful impact on the performance thermally of the cpu based upon the orientation of the cooler now this is not a knox of cooler i have on the table here but the idea was that if you have a cooler like this versus like this there was a difference purely from the level of contact or the type of contact with the heat pipes between the underlying dye under the IHS and the cooler itself ignoring all other variables like the case that was the theory so this is something that's unique to rise rise in 3000 especially because we're used to monolithic dyes from Intel and from AMD in the past but these asymmetrical multi chip approaches are still somewhat new the new rise in CPUs are arranged in a few different ways the first round of CPU is the more cut-down version of what became the 3950 X host a CPU die and an i/o die and the CPU die contains all of the cache the cores the i/o a I convinced the i/o as you would expect the 3950 X is different at host two dies for the cores in the cache and one for the IO still this modular approach means that all of them are asymmetrical to some extent but the 3950 axe is arguably a little bit less asymmetrical because it's got the full saturation of the substrate other than the ayodhya is kind of off alone so a 3,800 X versus 839 50 X will have 1 Delta of amount of silicon dyes on the board on the substrate that is because of this asymmetrical dye arrangement we've had viewers email us to ask if a Noctua and hu 14s would work better on a rise in cpu than the much larger NHD 15 entirely predicated on the idea that the heat pipe alignment would be more efficient and if you look at the NHD 15 its heat pipes are front to back whereas the U 14s are left to right and that's where that came from a few of the comments we found online also include these quote it's not whether you have one or two or ninety three towers but whether just one or two heat pipes end up being the ones the only ones covering the compute dies another quote single tower coolers usually have heat pipes running vertically when looking at the CPU as we see on the thumbnail of the quoted video which would mean at least two out of the four heat pipes would cover the chip flat if the single tower where to be rotated 90 degrees it would still be the same roughly two heat pipes directly contacting it but in this case or in the case of duel tower coolers because the heat pipes run horizontally with the two template CPU both triplets would be completely covered by the same heat pipes this actually is just a summation comment of what you saw in the the discussion another quote a final one this means most if not all cooling solutions available on the market presents you are ineffective since the dies are set in a symmetric non central fashion and that's why most people can't reach advertised boosts this is this last one is just an illustration of how reddit sees something and goes absolutely completely crazy certifiably insane with it these are two separate issues we don't really need to dig into it really any further than that but jumping from the base idea which does make sense of hey there's heat sources and asymmetrical locations maybe that impacts things to all coolers are ineffective is quite a large jump any let's trip maybe try and keep that in checking if everyone can just like kind of cool it a little bit on the sensationalism the dramatic responses that'd be great so you get the idea it's worth still trying this for ourselves since we don't really trust random Internet user data there's some of that out there but thermal testing is hard and we reran these tests two separate sets of times not two times with two sets of times for a total of a couple dozen verification passes between the three coolers and two CPUs tested Noctua is one of the few companies that does offer mounting hardware for both orientations so we use three Knox with coolers for this may attend small medium and large and Knox what also has come out publicly with its own statement dismissing the idea that there's difference here so we thought well we already have some data that we can source from them so we might as well just validate it relatedly the three models we're testing are the NH e 12 su 14s and NH d 15 the large one which is a dual Tower dual fan solution versus a single tower single fan for the others looking at the NHD 59 hu 14 s you'll see again it's left-to-right vs. front to back for the heat pipe orientation and then that's where the theories pops up so first of all if you're building in a case which almost everybody watching this will be this is all basically completely irrelevant because when you build in a case you're limitation is not going to be the die arrangement and the centralization of heat in a square on the motherboard your issue is going to be the panels of the case the airflow capabilities of the case the fans in the case the positioning of the fans in the case the positioning of other components in the case like the video card and things like that so a bad case can't be fixed by ultra optimized die to heatpipe positioning you're wasting your time even worrying about it a cooler that only fits in one orientation likely for GPU or RAM clearance reasons is obviously not usable in another orientation so not worth worrying about there either if you've got a cooler smashed against a glass panel on the top or double filters it's going to have a bad time again no reason to worry likewise using the top PCIe slot will smash a GPU right up against a lot of tower coolers which is about the worst possible configuration that's because clearly the GPU is a massive heat source that's potentially responsible for dumping 200 250 watts of heat into your case a lot of that is radiative heat off the backside of the GPU which would get sucked into the cooler the next aspect is that the GPU acts like a massive wall to airflow so the fan is going to need all the static pressure can get to fight the wall in front of it these things alone mean that in nearly every instance of a standard not inverted and non rotated ATX case a front to back configuration is likely the best solution probably not always but nearly always most people on the Internet in general unfortunately care more about seeing results than talking about how they are gotten different Seidel discussion so let's get into the results first and we'll talk about some of the testing parameters because these have to be very controlled and we're looking for something that might be a very small difference we'll talk conclusions after that well present the Noctua NHD 15 results on the 3800 acts first and then talk about the 3950 acts and some of the other coolers they use different dye configurations between the different CPUs so this allows us to better see if the heat distribution changes in a meaningful way when there is an asymmetrical dye layout under the IHS of course there is a change in the precise position of heat sources under the heat spreader obviously but the real question is whether that matters whether that impacts the teeth eye measurement when considering that solder and IHS multiple layers of metal in between and then Tim and then a heatsink are all part of the cooling solution with the 3800 acts and the D 15 looking at the first chart we were unable to measure a difference outside of our error we ran these tests six times or so and ended up with a 50 3.9 degree result for the rotated airflow pattern eg with air coming from the direction of the video card and then 52 point six degrees for the standard front to back orientation the video card here is a half-height passive GT 1030 in the bottom most compatible PCIe slot of the board so it's as out of the way as it can be and it's also short but it's obviously possible there's some very minor detriment from the video cards presence we couldn't find a reliable difference though between these results even with that consideration so these are functionally identical just for quick reference will highlight the VRM on the board as well erm thermals are largely unchanged from this rotation on an open bench all variables well controlled there's zero difference in the rotation of this cooler again the biggest change will come from adding a case and a bigger closer video card to the mix which would obviously benefit more from a front to back airflow path then bottom to top even if the risin 3000 eyes did care about this which they don't really seem to after considering the IIHS that solder and everything else in between the kulit it would still be cooler to orient it front to back in a standard case layout unless you're working with an inverted case or something like that but in a standard normal ATX case because of the positioning of the panels the presence of the panels the case fans the proximity of a big heat source that's also a wall that B the GPU all of these things would affect thermal is more than just the die layout would and die layout doesn't appear to affect thermals either which is why we test this on an open bench with none of those other variables or constraints just in case we tested the full 39 50 X chip with the same set of benchmarks this one still isn't tactically symmetrical but it's not missing a dial or a chip --let like the lower end CPUs are after six tests these were identical we were at 58.62 to the eight point nine degrees posting no difference outside of error vrm thermals were predictably higher for the 39 50 X rather than the 3800 X but that's not what we're comparing today you also shouldn't compare the temperature of the 3800 acts on the 39 50 X with these charts that's why they are isolated on different charts because they were tested with different voltages so that makes them incomparable you can't just look at one versus the other the 3950 hacks bend better so it's lower voltage so relative to running out the same voltage of the 3800 X it's less power than it would be anyway the two orientations the bottom of the top configuration ran warmer for the left side vrm and cooler for the top side vrm sensor which is completely predictable and normal that's just because of where the air is going and not for any other reason back to the 3800 X the Nhu 14s ran at 56 degrees TDI when oriented bottom to top or 55 point three degrees T die when oriented front to back this is again within test variance and error which is about plus or minus one degree here it doesn't matter what the died layout is in this testing we'll skip the vrm thermals for this one since they're irrelevant and also skill in the same way as last two tests but there on the screen if you really care about them finally the last chart the Knoxville NH u 12s on the thirty eight hundred acts can also be rotated so he tested that one as well in this instance the result was sixty point four degrees from the bottom top airflow and fifty nine point six degrees for the front back airflow our bottom top results are technically always between zero point seven ish and one point three degrees different always favoring the front back configuration although a pattern is forming and although it's tempting to say that the benefit is always for one configuration we can't call this causation it could be the somewhat distant but still close video card for example in the lower slot or something else for not accounting for it may be the RAM even though we position the fan on the same spot each time the RAM could also impact so we have to call these results functionally the same as for testing let's talk about this briefly and I get to the conclusion so the charts have some of the specs for the testing we set up and most of those that are relevant would be scenes things like the frequency and the voltage for the CPU that's very important things that need to be controlled that we did control positioning of bench in the room if you're not testing in a thermal chamber we've done that in the past too but if you're not testing on a thermal chamber and you're testing a normal room you need to be aware of the position of the bench the position of the ventilation and the ceiling and the status of the air conditioning or heating we never use heating here go figure we don't really need it and air conditioning is heavily controlled for our testing separately you need the ambient temperature at the inlet of the CPU cooler we monitor this second a second with a thermocouple we drop that data into the spreadsheet and then we produce delta T over ambient values in instances like this we don't always use delta T over ambient like when we know two devices are precisely controlled and can be directly compared with their based results but delta T over ambient is a simple subtraction of the ambient temperature from the temperature measured at the die level the component level this scales mostly linearly for the first at least couple degrees maybe five to ten or so and then stop scaling linearly after that but because we're never more than maybe one to three degrees different its this is the best way to do it and it controls things very well next one CPU core voltage must be fixed that's the biggest if you leave that auto everything you do is invalid CPU SRC voltage must be fixed all minor voltages should be fixed fan position and cooler must be controlled obviously for the D 15 we ensured that its fans were in the exact same vertical position for all the tests we drew an outline of the cliffs on the cooler so we could place them the same spot moving these fans even one inch in either direction can influence the results by a couple of degrees depending on where it's positioned on the test bench which direction it's in the same thermal compound has to use we use manually spread a synthetic compound we manually spread it so that we know how much is used in the past we've used things like graduated syringes as well to help control the amount although we have better methods these days blender 2.79 and the GM logo is used for the CPU load we monitor the CPU power consumption every second of the two actually every millisecond of the test and use that to determine that the power load is the same for all the tests and then the last few points here the render is held for approximately 23 minutes which is sufficient to reach steady state after just the first couple minutes and then we average from steady state we average a couple hundred rows of data cells of data to get our numbers that we report and then we do multiple test passes and average those so you're averaging thousands of cells of temperature data CPU temperature is is reported as TDI which is measured and hardware info that's probably about all we really need to talk about here so conclusions one of the earlier claims we saw that got a lot of traction on reddit was that the nhe of 14s would perform better than the NHD 15 due to the heat pipes alone and even in a perfect environment an open-air bench we cannot reproduce those claims and it's safe to say that they are wrong at this point we're not going to be able to reproduce those things ever so the only reason that might be true is if it's something to do with the positioning and the case and the obstructions to the cooler in that case that's the only reason that would ever matter not the dies so as we've shown tonight that's patently fall we were also unable to validate the claims in general that the heat sources particularly matter for overall like heat source versus heat pipe orientation and alignment doesn't really seem the matter on these three coolers now we haven't tested everything and maybe maybe something like a an open-loop block for example where you're dealing with jet plates that might matter a bit more but we haven't tested that so it's and also to be frank when you're talking up that level of cooling the delta is probably going to be basically irrelevant anyway even if it is a couple degrees because you're already so cold with one of those but with what we've tested today we feel fairly confident and stating that there's no meaningful difference and that you shouldn't worry about at all and you should continue about your day and worry only about things like how does it all fit within the case in the build as a whole not how does it align if the dies on your cpu you don't need to start drawing like the vinci code diagrams on your cpu to figure out how to place your cooler just put the damn cooler on the die and do it based on what your video card is and where your air intake is for the case probably front to back in nearly all instances unless you're working like a raven o2 well even that rotates it so unless you're working with a case that rotates it in a way where it's not facing the fans anymore then you can rotate the cooler that's it you'll get a bigger change from repositioning the fans in the case so thanks for watching subscribe for more come to storlock gamers XS net if you'd like to pick up some of our mod mats or shirts and it the way I phrase that made it sound like it's it's it's like a retail store but it's on the Internet you might be aware of it you can also go to patreon.com/scishow his access topside directly we'll see you all next timebakwin rise in 3000 launched there was reasonable speculation founded in basic physics that the asymmetrical dye arrangement of the CPUs with fewer chip 'lets could have implications for cooler performance the idea was that at the root of it a cooler whose heat pipes aligned to fully contact above the dye would perform better as opposed to one with two heat pipes sharing vertical contact with the dye we still see a lot of online commentary about this and some threads about which orientation of a cooler is best so we thought we'd bust a few of the myths that popped up but also do some testing on the base idea before that this video is brought to you by the msi gtx 1660 super gaming x the gtx 1660 super is a strong mid-range performer that's capable of 1440p gaming without issue and MSI's gaming experience uses a dual fan cooler with a more traditional two slot design for versatility it also has our RGB LEDs of course learn more at the link below this is pretty old news by now with much of the original discussion starting about two months ago but it's still alive and noctua recently revives this issue at the end of october knocked over knocked to a server at the end of october by stating that it believed there to be no meaningful impact on the performance thermally of the cpu based upon the orientation of the cooler now this is not a knox of cooler i have on the table here but the idea was that if you have a cooler like this versus like this there was a difference purely from the level of contact or the type of contact with the heat pipes between the underlying dye under the IHS and the cooler itself ignoring all other variables like the case that was the theory so this is something that's unique to rise rise in 3000 especially because we're used to monolithic dyes from Intel and from AMD in the past but these asymmetrical multi chip approaches are still somewhat new the new rise in CPUs are arranged in a few different ways the first round of CPU is the more cut-down version of what became the 3950 X host a CPU die and an i/o die and the CPU die contains all of the cache the cores the i/o a I convinced the i/o as you would expect the 3950 X is different at host two dies for the cores in the cache and one for the IO still this modular approach means that all of them are asymmetrical to some extent but the 3950 axe is arguably a little bit less asymmetrical because it's got the full saturation of the substrate other than the ayodhya is kind of off alone so a 3,800 X versus 839 50 X will have 1 Delta of amount of silicon dyes on the board on the substrate that is because of this asymmetrical dye arrangement we've had viewers email us to ask if a Noctua and hu 14s would work better on a rise in cpu than the much larger NHD 15 entirely predicated on the idea that the heat pipe alignment would be more efficient and if you look at the NHD 15 its heat pipes are front to back whereas the U 14s are left to right and that's where that came from a few of the comments we found online also include these quote it's not whether you have one or two or ninety three towers but whether just one or two heat pipes end up being the ones the only ones covering the compute dies another quote single tower coolers usually have heat pipes running vertically when looking at the CPU as we see on the thumbnail of the quoted video which would mean at least two out of the four heat pipes would cover the chip flat if the single tower where to be rotated 90 degrees it would still be the same roughly two heat pipes directly contacting it but in this case or in the case of duel tower coolers because the heat pipes run horizontally with the two template CPU both triplets would be completely covered by the same heat pipes this actually is just a summation comment of what you saw in the the discussion another quote a final one this means most if not all cooling solutions available on the market presents you are ineffective since the dies are set in a symmetric non central fashion and that's why most people can't reach advertised boosts this is this last one is just an illustration of how reddit sees something and goes absolutely completely crazy certifiably insane with it these are two separate issues we don't really need to dig into it really any further than that but jumping from the base idea which does make sense of hey there's heat sources and asymmetrical locations maybe that impacts things to all coolers are ineffective is quite a large jump any let's trip maybe try and keep that in checking if everyone can just like kind of cool it a little bit on the sensationalism the dramatic responses that'd be great so you get the idea it's worth still trying this for ourselves since we don't really trust random Internet user data there's some of that out there but thermal testing is hard and we reran these tests two separate sets of times not two times with two sets of times for a total of a couple dozen verification passes between the three coolers and two CPUs tested Noctua is one of the few companies that does offer mounting hardware for both orientations so we use three Knox with coolers for this may attend small medium and large and Knox what also has come out publicly with its own statement dismissing the idea that there's difference here so we thought well we already have some data that we can source from them so we might as well just validate it relatedly the three models we're testing are the NH e 12 su 14s and NH d 15 the large one which is a dual Tower dual fan solution versus a single tower single fan for the others looking at the NHD 59 hu 14 s you'll see again it's left-to-right vs. front to back for the heat pipe orientation and then that's where the theories pops up so first of all if you're building in a case which almost everybody watching this will be this is all basically completely irrelevant because when you build in a case you're limitation is not going to be the die arrangement and the centralization of heat in a square on the motherboard your issue is going to be the panels of the case the airflow capabilities of the case the fans in the case the positioning of the fans in the case the positioning of other components in the case like the video card and things like that so a bad case can't be fixed by ultra optimized die to heatpipe positioning you're wasting your time even worrying about it a cooler that only fits in one orientation likely for GPU or RAM clearance reasons is obviously not usable in another orientation so not worth worrying about there either if you've got a cooler smashed against a glass panel on the top or double filters it's going to have a bad time again no reason to worry likewise using the top PCIe slot will smash a GPU right up against a lot of tower coolers which is about the worst possible configuration that's because clearly the GPU is a massive heat source that's potentially responsible for dumping 200 250 watts of heat into your case a lot of that is radiative heat off the backside of the GPU which would get sucked into the cooler the next aspect is that the GPU acts like a massive wall to airflow so the fan is going to need all the static pressure can get to fight the wall in front of it these things alone mean that in nearly every instance of a standard not inverted and non rotated ATX case a front to back configuration is likely the best solution probably not always but nearly always most people on the Internet in general unfortunately care more about seeing results than talking about how they are gotten different Seidel discussion so let's get into the results first and we'll talk about some of the testing parameters because these have to be very controlled and we're looking for something that might be a very small difference we'll talk conclusions after that well present the Noctua NHD 15 results on the 3800 acts first and then talk about the 3950 acts and some of the other coolers they use different dye configurations between the different CPUs so this allows us to better see if the heat distribution changes in a meaningful way when there is an asymmetrical dye layout under the IHS of course there is a change in the precise position of heat sources under the heat spreader obviously but the real question is whether that matters whether that impacts the teeth eye measurement when considering that solder and IHS multiple layers of metal in between and then Tim and then a heatsink are all part of the cooling solution with the 3800 acts and the D 15 looking at the first chart we were unable to measure a difference outside of our error we ran these tests six times or so and ended up with a 50 3.9 degree result for the rotated airflow pattern eg with air coming from the direction of the video card and then 52 point six degrees for the standard front to back orientation the video card here is a half-height passive GT 1030 in the bottom most compatible PCIe slot of the board so it's as out of the way as it can be and it's also short but it's obviously possible there's some very minor detriment from the video cards presence we couldn't find a reliable difference though between these results even with that consideration so these are functionally identical just for quick reference will highlight the VRM on the board as well erm thermals are largely unchanged from this rotation on an open bench all variables well controlled there's zero difference in the rotation of this cooler again the biggest change will come from adding a case and a bigger closer video card to the mix which would obviously benefit more from a front to back airflow path then bottom to top even if the risin 3000 eyes did care about this which they don't really seem to after considering the IIHS that solder and everything else in between the kulit it would still be cooler to orient it front to back in a standard case layout unless you're working with an inverted case or something like that but in a standard normal ATX case because of the positioning of the panels the presence of the panels the case fans the proximity of a big heat source that's also a wall that B the GPU all of these things would affect thermal is more than just the die layout would and die layout doesn't appear to affect thermals either which is why we test this on an open bench with none of those other variables or constraints just in case we tested the full 39 50 X chip with the same set of benchmarks this one still isn't tactically symmetrical but it's not missing a dial or a chip --let like the lower end CPUs are after six tests these were identical we were at 58.62 to the eight point nine degrees posting no difference outside of error vrm thermals were predictably higher for the 39 50 X rather than the 3800 X but that's not what we're comparing today you also shouldn't compare the temperature of the 3800 acts on the 39 50 X with these charts that's why they are isolated on different charts because they were tested with different voltages so that makes them incomparable you can't just look at one versus the other the 3950 hacks bend better so it's lower voltage so relative to running out the same voltage of the 3800 X it's less power than it would be anyway the two orientations the bottom of the top configuration ran warmer for the left side vrm and cooler for the top side vrm sensor which is completely predictable and normal that's just because of where the air is going and not for any other reason back to the 3800 X the Nhu 14s ran at 56 degrees TDI when oriented bottom to top or 55 point three degrees T die when oriented front to back this is again within test variance and error which is about plus or minus one degree here it doesn't matter what the died layout is in this testing we'll skip the vrm thermals for this one since they're irrelevant and also skill in the same way as last two tests but there on the screen if you really care about them finally the last chart the Knoxville NH u 12s on the thirty eight hundred acts can also be rotated so he tested that one as well in this instance the result was sixty point four degrees from the bottom top airflow and fifty nine point six degrees for the front back airflow our bottom top results are technically always between zero point seven ish and one point three degrees different always favoring the front back configuration although a pattern is forming and although it's tempting to say that the benefit is always for one configuration we can't call this causation it could be the somewhat distant but still close video card for example in the lower slot or something else for not accounting for it may be the RAM even though we position the fan on the same spot each time the RAM could also impact so we have to call these results functionally the same as for testing let's talk about this briefly and I get to the conclusion so the charts have some of the specs for the testing we set up and most of those that are relevant would be scenes things like the frequency and the voltage for the CPU that's very important things that need to be controlled that we did control positioning of bench in the room if you're not testing in a thermal chamber we've done that in the past too but if you're not testing on a thermal chamber and you're testing a normal room you need to be aware of the position of the bench the position of the ventilation and the ceiling and the status of the air conditioning or heating we never use heating here go figure we don't really need it and air conditioning is heavily controlled for our testing separately you need the ambient temperature at the inlet of the CPU cooler we monitor this second a second with a thermocouple we drop that data into the spreadsheet and then we produce delta T over ambient values in instances like this we don't always use delta T over ambient like when we know two devices are precisely controlled and can be directly compared with their based results but delta T over ambient is a simple subtraction of the ambient temperature from the temperature measured at the die level the component level this scales mostly linearly for the first at least couple degrees maybe five to ten or so and then stop scaling linearly after that but because we're never more than maybe one to three degrees different its this is the best way to do it and it controls things very well next one CPU core voltage must be fixed that's the biggest if you leave that auto everything you do is invalid CPU SRC voltage must be fixed all minor voltages should be fixed fan position and cooler must be controlled obviously for the D 15 we ensured that its fans were in the exact same vertical position for all the tests we drew an outline of the cliffs on the cooler so we could place them the same spot moving these fans even one inch in either direction can influence the results by a couple of degrees depending on where it's positioned on the test bench which direction it's in the same thermal compound has to use we use manually spread a synthetic compound we manually spread it so that we know how much is used in the past we've used things like graduated syringes as well to help control the amount although we have better methods these days blender 2.79 and the GM logo is used for the CPU load we monitor the CPU power consumption every second of the two actually every millisecond of the test and use that to determine that the power load is the same for all the tests and then the last few points here the render is held for approximately 23 minutes which is sufficient to reach steady state after just the first couple minutes and then we average from steady state we average a couple hundred rows of data cells of data to get our numbers that we report and then we do multiple test passes and average those so you're averaging thousands of cells of temperature data CPU temperature is is reported as TDI which is measured and hardware info that's probably about all we really need to talk about here so conclusions one of the earlier claims we saw that got a lot of traction on reddit was that the nhe of 14s would perform better than the NHD 15 due to the heat pipes alone and even in a perfect environment an open-air bench we cannot reproduce those claims and it's safe to say that they are wrong at this point we're not going to be able to reproduce those things ever so the only reason that might be true is if it's something to do with the positioning and the case and the obstructions to the cooler in that case that's the only reason that would ever matter not the dies so as we've shown tonight that's patently fall we were also unable to validate the claims in general that the heat sources particularly matter for overall like heat source versus heat pipe orientation and alignment doesn't really seem the matter on these three coolers now we haven't tested everything and maybe maybe something like a an open-loop block for example where you're dealing with jet plates that might matter a bit more but we haven't tested that so it's and also to be frank when you're talking up that level of cooling the delta is probably going to be basically irrelevant anyway even if it is a couple degrees because you're already so cold with one of those but with what we've tested today we feel fairly confident and stating that there's no meaningful difference and that you shouldn't worry about at all and you should continue about your day and worry only about things like how does it all fit within the case in the build as a whole not how does it align if the dies on your cpu you don't need to start drawing like the vinci code diagrams on your cpu to figure out how to place your cooler just put the damn cooler on the die and do it based on what your video card is and where your air intake is for the case probably front to back in nearly all instances unless you're working like a raven o2 well even that rotates it so unless you're working with a case that rotates it in a way where it's not facing the fans anymore then you can rotate the cooler that's it you'll get a bigger change from repositioning the fans in the case so thanks for watching subscribe for more come to storlock gamers XS net if you'd like to pick up some of our mod mats or shirts and it the way I phrase that made it sound like it's it's it's like a retail store but it's on the Internet you might be aware of it you can also go to patreon.com/scishow his access topside directly we'll see you all next time\n"