Cooling CPUs will never be the same after using this!

I forgot I even had these this is the the cryo sheets from thermal Grizzly you know devours company and basically it's designed to replace thermal paste I'm used to using these big old giant just of thermal paste so I figured today we would just kind of give these a shot and see if uh well how well they perform I guess [Music] the air 903 Max mid tower case from montec features a 51 porosity ventilation rate pre-installed HP 140 millimeter argb fans and 21 built-in RGB modes for a premium feel without the premium price not interested in RGB then the air 903 base is just for you with its clean aesthetic and incredible component support at a price that won't have you crying to see the full list of air 903 specs from montec follow the link in the description below so today I'll be putting these on my 13900k which is my test bench and it's going to basically be kind of going up against the Kingpin cooling so kpx for the Kingpin extreme cooling uh thermal paste is what I pretty much like exclusively use and kind of like here's some proof of it where I've got you know lots of it each one of these tubes will go a long way I use it on gpus I use it on CPUs I use it on memory modules I use it everywhere but the benefit to using like a graphite graphene sheet is the mess because remember this is all a liquid right so it comes out goopy like you just saw right there when I took the cap off it's like but anyway it gets messy you put the cooler down you tighten it down it squishes out the sides it go over the sides of the CPU and you got to clean it it gets all ugly but the concept with these is that you'll put them on they'll still feel the feel the gaps the cryo sheet is basically designed in a z-axis where the cooling molecules I don't know the science behind it basically or designed in a z-axis format to help create a better thermal transfer I didn't know there was grain or direction to cooling but apparently that's what they advertise so that's what I'm saying but I also noticed these are different sheets so there's a 24 mil by 12 mil there's a 33 by 33 and a 38 by 38 so obviously these are designed for different size CPUs so 24 by 12 and we take a look at this it's actually intended for like delighted CPUs this is my d-lated 3900k here I won't be using that today because I do not have well I do have technically um Singularity computers direct die GPU or CPU block but that's not what I'm using today but you can see the size of the sheet right there so that's obviously intended to be used in like a a direct dye cooling situation which um I guess again would be really nice to keep it from getting all over the place in there that square right there might be kind of hard to see that's clearly going to be for like AMD let me grab one of their CPUs so if we compare the size of the sheet this is a like an old school ryzen you can see it doesn't fully cover so I'm not I think the 38 by 38 would work for this try not to touch it okay stay in between the sheets thank you that's what I say my wife just kidding yeah that would work for am5 pretty good it would also work for old older Intel as you can see fully covers that I have no idea how fragile this stuff is and as you can see I have a finite so let's look at the 38 by 38 yeah that's a big boy right there yeah that'll clearly cover it but you're gonna have it hanging off the sides as well maybe it was just easier for him to make it in a square though it's got better applications but as you can see that will fully cover but it will dangle off the side it's not going to matter but this would also work with like x99 CPUs and stuff which are a bigger IHS um xenons and all that but this won't work with threadripper threadripper is like way too big so this is my rig we're going to use the 3900k on a 360 a nucleus AIO from EK water blocks non-rgb very basic nothing crazy about it GPU doesn't matter this is just on here for display it's just a 3080 you know no big deal yeah this one's not the ti no it's not the ti no The Edge right here would be shiny if it was it's not shiny enough yeah it's it's pleb one anyway a couple things the fans are static at 75 percent now I'm trying to control as many variables as possible however I do not have a locked voltage but I do have a locked Max voltage now I am running asus's AI Optimizer on this so it's not going to be as Apples to Apples as I'd like meaning if I'd gone in a locked down all the perimeters but I'm kind of approaching this in the if you're a regular guy Regular Joe that most people do not go in and lock down their frequencies and lock down their uh voltages and such but they'll just go okay I'm going to go from thermal paste to this so what happens that's what I'm kind of looking at now I want to point out that it's probably going to be extremely consistent but I the AI Optimizer is actually performing really really well so you can see right now the system while idling 1.462 volts 1.45 that's a lot of voltage but look at the temps on the the cores right now so this thermal or this graphene pad has a lot of work cut out for it because I am idling in the 20s on the P cores and obviously in the 20s on the E chord it's also at 6.1 gigahertz single core and it will go all core 5.6 but look the rest of the cores are chilling at four equals or at 4.6 and then we're at 4.7 4.8 4.9 or it's gonna be 5.7 5.8 5.9 and then 6.1 on the top three cores but watch what happens when I go ahead and put this under load our most my most recent run was a 4000 or 40 129 watch out through the voltage it's going to drop 1.18 volts this and it's at 73 74 in the package cores are in this low 70s high 60s 5.6 gigahertz all core could push that higher if I wanted this is a magic chip very rarely do I get one that's actually really good this one is really good that just ran a 40 251 I just went higher than where we just were so yeah what we need to do right now though on our Baseline is I'm going to start the test I'm going to run a 10 minute loop on this and then what we're going to compare here is one where are Max temperatures land but I also have to pay attention during the test to see what's happening with the voltage because if the temp comes down but the voltage came down that means it was a result of voltage and not the graphene pad this is why locking down the voltage would be great but I can't go in and set a static voltage because What's Happening Here is right now while the C the CPU is not under load you can see my cores are boosting up pretty high right now 5.8 right now 6.1 if I were to go in and lock the voltage at 1.18 or whatever it's doing under load conditions then obviously it will crash it won't even boot so I have to allow it to have that fluctuation just because the way the CPU is set for our test bench but that's a this is a pretty good real world scenario test where it's like what happens if you take this in a real world functioning system put this in and then see what happens it may not be the best Apples to Apples comparison but it is still a result all right so I'm going to start the test right here 10 minutes multi-core start after it starts I'm going to clear my min max clear min max there we go packages at 77.78 right now we're hitting like 78 on a couple of cores the reason why we have to let this run is because we have to let the coolant get to equilibrium that's what we're really going for here is what happens when the cool inside its Max temperature but realistically I care less about the score and I care more about what's happening with our Max temperatures like our package was about to hit 7 or 80 for a second there oh no it's hitting 80 at five six all core oh it just crashed all right so because of that crash we did a slight tweak I basically bumped the LLC up one level so now we're hitting 1.2 volts which is fine um but I bumped the fans from 75 up to 100 just to keep the temps I don't know I guess I could have let them get hotter but I'm trying to keep it as stable as possible but anyway you can see right now slightly More Voltage 1.203 towards the end of the test 1.197 but towards the end of the there it is 1.2 204 and then it will go up in the in between the runs when it restarts so now we're off and running still five six gigahertz all core four three on the e-core um e-cores are at 65c hottest core is at 68. our highest score right now on the P cores is um P cores Max says 79 but the individual core doesn't show it anyway shows the max of 79. we can go from there so all right we're off and running we're already past where we were we haven't crashed in terms of the score that last one was a forty thousand one ninety seven so we didn't lose any performance with it all right let's Let It Go and uh get the data not the most scientific way to do it but I want to show you what the ambient air temp is in here it's 19.2 yeah this couple is intends to be used while touching something but it's it's still works pretty good this way watch if I touch it look at that so but now it'll take time to bleed off that heat if I wave it around I'm gonna come back down so 19.2 is what the current Ambient Air Temperature is so that way you guys can see you know and there's another point of data air temp ambient temp obviously matters with cooling your system all right so we are about to finish 20 seconds left in the test it has not crashed the max package temp achieved was 82c Max cores though was uh between well 80 was the max core right there and then everything else is in the upper 70s I think that one hit 80 right there as well for a second voltage you can see it's really consistent we'll show you as this run completes because it just finished the 10 minute this is the last run 1.198 1.197 it will go up to about 1.205 towards the end of the test anyway the test just completed and that last was a forty thousand to elevens extremely consistent it stayed around that 40 200 and some change marked the entire time um I need to cool off this temp the sensor a little bit again you know I have I have an actual better tool for this okay there it is 19.2 so it didn't change obviously if I blow one in right now that's not gonna do any good because I'm obviously full of hot air we'll use the correct tool here the manometer then a manometer all right moving on now the system has cooled down a little bit I want to let it idle here to bring the coolant temp down some so that's part of the problem with thermal paste as soon as I took it off it just started like everywhere I mean this is clearly one of the cons to thermal paste is the cleanup so it's not just a potential thermal benefit depending on the thermal paste you're running there's no possible way that thermal paste could get down inside the pins if you were to take the CPU off you got all this hanging over the side so there's a lot of Pros to the idea of this as long as it performs well one of the other major Pros is going to be the fact that it won't dry out over time whereas thermal paste can and will because it's especially if it's an alcohol-based alcohol evaporates right and so that's why eventually the thermal paste sometimes you see gets super hard and I'm not going to clean that what's down in there around the sides right now because I don't want to take the CPU out in between these tests right now but I am making sure that I get every single bit of the thermal paste off the CPU because we don't want to contaminate our tests by getting any of that residue on there whatever hangs over the side isn't going to be a big deal because that's not touching anything important anyway but same goes for this the block here I've got to get every single bit of this residue off which is going to be quite a lot of it now the whole point of thermal paste and as I move this around you can see there's a lot of scratches in all the metal like especially these the copper blocks and stuff they're really soft but even on the IHS there's a there's a significant amount of um microscopic Peaks and valleys in the metal and all thermal paste does is fill those Peaks and valleys to get better contact between the two the two pieces of metal for better heat transfer so that's all that the pad has to do as well but I gotta get my back plate put back on there actually you know what I forgot how much I love this test bench because it just pressure fits in there so I'm wondering how much this thing's going to slide around when I go to install it hopefully not too much see if I can do this without touching it because it's not going to like stick down or anything like that in fact it's got a little bit of a curve to it it's also fairly fragile so you could rip it pretty easily that's lined up a little bit of surface tension right now um kind of making it sort of stay in place as long as I'm able to put the cooler down and not be twisting it all around we should be good don't let it move because the way this video is going to end is we are going to be taking it off and seeing what happened to it like did it rip um apparently according to Phil who's researched this particular topic a lot because of like the Ally and all that and the steam deck and whatnot a lot of these companies are using these kinds of pads now but that they liquefy as they get hot but then in their more of like a room temp state they're very solid the question I have is is this a one-time use that's that's what I want to like once you take the cooler off kind of like what the AMD gpus is it done or can you use it again what happens with the AMD gpus is they tear however the AMD gpus are also a chiplet design so there's different Peaks and valleys in there that get compressed whereas this is just a single rectangle giving us uh our IHS so a little bit of a different scenario there so we kicked on our HVAC to make sure the temperature stays close is this getting hotter throughout the day so at 19.4 right now so we just want to keep that in mind I'm going to hang that over the edge so air can flow through it okay thermal pad is in there our initial temperatures are all pretty much the same our package is idling in the high 20s low 30s depending on the load that's kind of bouncing around but I want to see what the initial startup temperature is that'll tell me right away if good or bad is going to precede what happens next if I started and it's higher off the bat then that's going to be a bad thing boom package temp 67 72 73 exactly where it was with the thermal paste 74 oh 76. I feel like I went to 76 a little bit quicker initially we're at five six all court on the P cores 4 3 all core and we're at the exact same voltage that's good I was hoping that there wouldn't be any variable change in there when it comes to the voltage and stuff so this now gives me a realistic a b comparison even though the variables were not locked down like they should forty thousand Two Eleven was the very last run of the 10 minute Mark last time we're at forty thousand one seventy margin of error honestly but if it hits lower in like the Forty thousand one hundred consistently versus the 200 there's a few different reasons that could be um but we'll talk about that at the end all right 17 seconds left in the test 19.1 C air temp so that's Air temps not going to be a factor in this check this out we are actually slightly higher on the voltage at the start of the next run see if it'll start before this test ends 1.204 at the start then it drops to 1.198 then 112 you know 1201 1197 1196 I'm only bringing that up because of the fact that um we are on average about 1C hotter still the max the max package was 80. so that's exactly the same to be honest our wattage draw is still the same at 268 Watts that's nuts like that it's actually 268 watt draw giving us this performance but look at the score it's actually slightly better 40 000 235 versus 211 margin of error that means nothing but the point is we lost nothing that's a good thing right so let's talk about this pad real quick before we shut it down as it's cooling down now so it has now performed on par with I already put it away it's the kpx which I use everywhere there's a positive and a negative to this one the instructions actually say to cut it down to size obviously I didn't cut it down to size but as you see that didn't affect anything negatively here kpx like all thermal paste is messy you have to get good mounting pressure and it squishes out the side and you saw how it goes all down the side of the CPU and if you get really too liberal with it it as you take it off it can goop like you I showed with the cooler down into the socket and that's a bad thing now it's all hard to clean you can actually just kind of douse the crap out of it with isopropyl alcohol that'll break down the kit the thermal paste and you can flush it out pretty easily without ever having to touch it because those pins are very delicate the graphene pad by the way even though it's a solid pad it does not liquefy it's not that thermal pad that turns into a liquid like a phase change pad this is a graphene solid pad the higher the mounting pressure the better it's also conductive whereas kpx and other thermal paste are not so even though the thermal paste can get everywhere as long as it doesn't cause physical damage in the socket if you get it down in there and you just went nuts with it if this somehow were to when you're putting it on sort of slip and maybe touch something which would be actually kind of hard where the socket and stuff is it's just worth pointing out it is conductive all right so now that we our scores were extremely consistent um let's shut it down and see what happens when you take it off because if as long as it doesn't get damaged I would have to assume it's possible to reuse it and that's what I want to see right now if I take it off right now it doesn't look damaged in any way I'm going to put my cooler back on it and then run the test again and see what happens because it's 25 bucks for that one pad you know that's a one-time use for most people that's not going to be a problem because they put their computer together and they leave it until they have to do something and usually the only reason they take off their cooler is either to reapply thermal paste because it started drying out and the temps started going up or they've upgraded their cooler or something if you have to pay 25 bucks every time you go to reuse it like I'm gonna be sad right now if it's not reusable because I only have three of this size hey look at that stayed in one piece check that out you can see the contact pressure of where the CPU was you can also see as a slight tear right there so I think as long as I can get this back down without that tear being in the way should be okay because it they really do want me to trim that down to size but obviously I didn't and I'm not gonna so let me get this look at that there's no cleanup you can even see the imprint of where the CPU was which is kind of nice as I'm reinstalling this too there's one thing I want to point out and that's going to be that obviously the easiest installation orientation for this is going to be having your system flat like this because if it's sliding all around with it's vertical try and put the cooler together like that that's going to be a it's not going to go too well so you're obviously going to want if you're installing this in a motherboard maybe that's already in a case you're clearly going to want to uh lay it down on its backside like this otherwise you're going to have a heck of a time trying to get it to not slip out of place so initial attempts on the startup look uh pretty much exactly where I expect them to be oh and we are up one degree a little over one degree in ambient right now 73-74 core clocks are exactly where they were before wattage draws exactly where it was 266. there's a 40 200 so it's only 35 points below and that again that's always going to fluctuate like there's always going to be plus or minus like 20 30 points in this test and yeah I'm also aware cinebench r24 is out we might do a video about it but um this test is still plenty good for the AVX construction set to put heat into the CPU and test things like thermals right here so it already has a better proposition value and cost by the fact that you I've at least reused it once and we think that little tear probably happened because I didn't trim the sides so it wasn't Collide against the CPU as it hangs over the side it has to create that radius right so it tore like that I've seen these types of pads I've always seen the phase change ones though that have like a finite life to them still and can potentially dry out because they are liquid based this is graphene it is it will this will last as long as your system is together or until you physically damage it it will not let if you ran the if I ran the system for five ten years that pad will be as good as it is today in five or ten years that's the craziest part about it it will not dry out it will not degrade now if I could use it twice that brings the cost of it down to 12.50 but still that's way more expensive than because it's about five dollars for a one use tube of thermal paste so you'd have to use it five times to get your full money worth but most people are not tearing their systems apart five times I mean we do around here because we test product but most people don't have to do it except once every couple years okay well there you go um I'll be honest I didn't realize I had these until I was going through the shelf and I'm like when did we get cryo sheets I know he sent these to me a long time ago so I don't know how long they've been out I know it's not a new product by any means but it's uh I'm glad I took a look at it because I feel like I'm gonna stock up on these and start using these on our test benches and stuff just because of how often we change out CPUs I'm tired of cleaning thermal paste that gets everywhere and the cool thing is this means at the end of the day I won't be putting CPUs away that look like this or they're still thermal paste and the crevices and crevasses of it because I'm too lazy to clean it out like every single one of our our CPUs have thermal paste gunk in there that won't be the case with these look like I know this one actually cleaned pretty good but whatever that's awesome all right guys thanks for watching and as always we'll see you in the next one

Channel: JayzTwoCents

Views: 1,088,272

Keywords: thermal paste, best thermal paste, how to cool a cpu, best cpu cooler, best cpu thermal paste, best paste for cpu, graphene pad, graphene cooling pad, cpu graphine, kryosheet, what is kryosheet, thermal grizzly

Id: pTRk52LpRco

Length: 22min 34sec (1354 seconds)

Published: Tue Oct 03 2023