i see the little boss and who really runs the shop is behind you working hard lol

@Xenoray1

Ай бұрын

my humor is so dark, i cant make a joke here.

@lonewolfrcs1331

Ай бұрын

I have a quick question and I can’t seem to find a straight answer to I have a 360 Corsair aio will it hurt to run it a 100 percent pump speed at all times ?

@OneShotCS

Ай бұрын

@@lonewolfrcs1331 I have my ek 360 running 100% all times and it's fine

@cloudyyy-._.-999

Ай бұрын

@@lonewolfrcs1331 why're you trying to run 100% all the time?

@GrySgtBubba

Ай бұрын

Lmao Spot on!

The "veins" You see are created from the suction force that's generated when you pull the cooler off of the CPU. If you were to use a clear piece of plexi to push down on the CPU with that same grease you would see no veins until you pulled it back off again.

@craigjamroz7585

Ай бұрын

Exactly was thinking the same thing 😄👍

@UncleButterworth

Ай бұрын

More than likely.

@fitztech

Ай бұрын

I think Jay is starting to lose his mind tbh

@fabiokinolli

Ай бұрын

The main point was to show how prominent they are compared to your typical thermal paste, normal thermal paste will show veins nowhere near to the degree this one did

@LockDots84

Ай бұрын

​@@fabiokinolli You're not wrong about the veins being more prominent because of the grease being thinner but I don't think that was the main point he was trying to make. He mentioned that he thinks they show up based on high/low spots but I believe it's purely due to suction/vacuum force when removing.

"I'm glad she's wearing headphones back there" lol

@kaseyboles30

Ай бұрын

I seriously doubt anyone in Jay's life is unaware his humor is sometimes that of a 14/yo boy.

@leagueaddict8357

Ай бұрын

Honestly, even if it is sound canceling she probably can still hear some of it you can also see this when she turns her head, she is picking up some of the sounds.

@watercannonscollaboration2281

Ай бұрын

JayzTwoNSFW

@terrynt

Ай бұрын

headphones != sound proof

@awwtv2603

Ай бұрын

She will see it on youtube.

Jay, that paste will only conduct heat well after it dries. You must "cook it" first in situ (meaning already applied and everything mounted). It must resemble those flakes that came of the tube when you tried to apply it the first time. I have used it before while repairing coffee machines. It is used as interface between the resistance and the boiler. So, you should apply it and either let it dry naturally (which can take months) or crank that CPU for a few days. One important thing is that, if you remove the cooler, you must clean the paste and reapply otherwise the cooler will not seat properly and the heat will not transfer as effectively because the dried paste breaks very easily.

@PhonicallyPsychotic

Ай бұрын

I wonder if delidding and baking it onto the lid would work ? You would probably want to leave it a little tacky.

@timhartherz5652

Ай бұрын

I guess on paper it has a better conductivity, but only in the right use case which this is not. A CPU us not supposed to get as to the point where this paste is supposed to be operated at. Maybe it would be good for some MOSFETs but little more. That's a nope for computer application.

@kloroformd

Ай бұрын

@@timhartherz5652 True, but also at least it'll be non-conductive at Intel-spec in anything they made the past decade. Or Apple spec from anything in the past 20 years.

@timhartherz5652

Ай бұрын

@@kloroformd "Within specs" for these companies mean "will survive the mandatory warranty period" :-D

@kloroformd

Ай бұрын

@@timhartherz5652 if not, they'll polish a used item for ya. There's probably an arbitration clause embedded in the firmware for those that want to argue.

Der8auer had a video some time ago, I believe with Steve from GN were they discussed thermal conductivity and why thermal grizzly and others have gone away from the W/mK values. There's no Industry Standard on how to measure that. Basically, the hotter the heater the thermal interface is tested with, the higher the thermal conductivity in W/mK

@oddsketch

Ай бұрын

facts

@mastersprite2783

Ай бұрын

Is it really Watts per meter Kelvin? As an engineer, when I see units written like that my brain goes milli-Kelvin.

@bythelee

29 күн бұрын

The devil is in the details. The problem is that the material with the headline conductivity (be it silver, graphene, or Boron Nitride) is usually only ONE of the ingredients making up the paste. The ratio of high conductivity material to the carrier matrix makes it a lottery whether the compound will be good or bad at heat transfer. For example, a TIM with 1% silver and 99% insulating mush will be next to useless, where something with 99% silver and 1% mush will be almost as conductive as pure silver. Kinda obvious and intuitive, but finding out the % ingredients in any TIM is next to impossible, because they are "closely guarded trade secrets". Field testing to determine what works and what does not, is the only way to sort wheat from chaff. And we have to ignore the headline conductivity of the buzz-making ingredient because it gets swamped by all the other stuff going on. The burger analogy (with all the layers) further subdivides within the paste layer, where heat is mostly moving through the "solvent" (the material that glues all the particles together) but can use the high conductivity particles to "short circuit" across the higher resistance matrix. Again, the percentage of high conductivity material greatly influences the conductivity of the composite, including whether the particles of the high conductivity material actually touch each other (allowing a more continuous path of highly conductive material through the compound). It is never that simple, though, with factors like surface roughness of the particles (and even whether the particle surface has oxidised, adding a layer of resistance to the surface, etc) all affecting the score. The only effective way to measure whether the TIM works, is a real-life test like this one. Good developers will experiment with their product ingredients to find an "optimal" composition. Which may or may not include a necessary process of installation (like "cooking" the Boron Nitride paste before it works properly). Graphene is an outlier here, since it is "pure carbon". But it equally misleads, because the very high conductivity of a graphene sheet is ONLY "in the plane of the graphene". Unlike most materials which offer the same value of conductivity in all directions, graphene is extremely directional - because a graphene sheet is only one atom thick. What is marketed as "graphene" is actually a very thick stack of multiple graphene sheets to form a paper. When multiple layers of graphene are stacked together to make a TIM, then heat has to travel from layer to layer, and the conductivity of that assembly is very much less than the in-plane number (I'd estimate only about 10% of the in-plane value). So why does graphene paper work well as as TIM? Because, heat enters from the heat spreader at a few very local hot spots. The graphene can then spread that heat about the in-plane area extremely well, using that high conductivity. The less effective layer-to-layer transfer is then mitigated, because it can use ALL of the contact area for that transfer. With most (non-directional conductivity) TIMs, there is no diffusion of heat within the layer - heat spreads from the source hotspot in a spherical manner, which barely increases the area of transfer at all, before the heat is already passed through the thickness into the cooler.

@bythelee

29 күн бұрын

@@mastersprite2783 Lol. Yeah, you're technically right about "mK". The definitive proof otherwise is that if you do Dimensional Analysis of what the number is made up of, you get "heat energy per unit of distance per degree of temperature". If it was indeed milliKelvin, then the distance unit would be missing.

@batoregokomin

29 күн бұрын

@@bythelee actually it's not how graphene thermal pad is made, at least not Kryosheet from derbauer. He explained it in one of the video. The graphene sheets are stacked and baked together into a thick block and then it's sliced from the sides into a thin sheets, so the graphene planes are actually perpendicular to the surface of cooler coldplate and CPU.

This is interesting, but would be nice if jay would leave the paste dry out as is mentioned in manual. You should do the 10 minutes test, but you only tested it in wet condition. " The paste is aqueous, so the water carrier must evaporate before the compound becomes an electrical insulator. Paste bridging a temperature sensor’s lead wires will affect the sensor’s measurements while the paste remains wet. Dry the paste by allowing it to set overnight, or to save time, heat it to a temperature of 90°C for ten minutes. The paste will dry completely, and temperature measurements will return to normal upon first use of the hotend. Avoid exceeding 100°C while drying, to keep the water from boiling and ejecting the paste with steam."

@Revan_7even

28 күн бұрын

RTFM Jay, not long enough with nowhere for the water to go. This isn't as bad, but don't be Linus with Billet Labs because of impatience and not wanting to go through the hassle.

@GlueFactoryBJJ

28 күн бұрын

Not to mention that, IMO, he put WAY too much on the CPU. The best conductivity would be metal to metal, but since the metals aren't perfectly smooth, there are, as Jay noted, "voids" and non-contact points (due to the CPU/cooler not being perfectly flat) and there needs to be a minimal conductive paste bridging those gaps. As you noted, water isn't the best temperature conductor (only 0.6089 w/mK - Wikipedia), so it would need to be evaporated before the thermal paste would perform its best. So, yeah, I'd like to see this test repeated. Especially, since it doesn't seem to be being used in an optimum manner in this test. Anyway, my "$.02"...

@arthurmoore9488

3 күн бұрын

@@Revan_7even Good news is he's not naming companies. This is an example of how preconceived bias can result in a flawed test. In this case it's a reasonable, "The product was not designed for this application."

Just an FYI, this shit once set in, creates almost permanent bond between IHS and a cooler.

@ajslim79

Ай бұрын

perfect for AM4 CPUs it seems :P

@IcecalGamer

Ай бұрын

5 minutes test at 90C is not nearly enough to cure it. Not a representative test. Edit: It wasn't used in design specs :/

@PolskiJaszczomb

Ай бұрын

@@IcecalGamer It dries naturally over time. Not that the IHS sits at 90C when the CPU is at such temperature, you need ZERO cooling for the IHS to reach this temperature. Else it's impossible to even come close to 70C.

@Timmy_420

Ай бұрын

@@PolskiJaszczomb Not really, my miniPC is sitting at 85C right now while only on the internet with only around 22% of the CPU and 39% of the DDR5 memory in use. And the iGPU is sitting at 59-63C. So much for your predictions but I suppose this is true with at least custom/self-built PCs because when mine was up and running, it was usually around 50-60C while online and even gaming

@PolskiJaszczomb

Ай бұрын

@@Timmy_420 Do you realise, that the iGPU is in the silicon die, that also houses main cores? You'd be at >constant< 80+C with an IHS of 60C.

What does your dad do all day ? She said : He watch a picture of a kitchen draw on his monitor.

@Rockr69747

28 күн бұрын

more like a knook

@noisullI_dnarG

28 күн бұрын

Not a kitchen

@thomgizziz

28 күн бұрын

@@Rockr69747 nook?

Slice Engineering does 3D printing stuff BUT in that application there is usually not the pressure that PC coolers have. the layer might be thicker and doesn't spread out as much. layer thickness is much more important that the heat condusctivity. we are all good off using the products that are application appropriate.

The calculations for W/mK depends on a lot of factors and are easy to fudge to get a number that looks great but doesn't apply to PCs. Things like the temperature difference between the heat source and cooler out the thickness of the thermal interface material can make it very wildly. This stuff may work extremely well in its intended application but less so in PC cooling. In any case, glad you tested it, it's always nice to see things tested in different ways by different people.

@arthurmoore9488

3 күн бұрын

As others have noted, the temperature it's tested at is a key factor. Given this is designed for 3D printing, it was likely tested at 200+C. So, twice the temp!

Hey Jay, Hope things are going well healthwise! Praying for you to get back in full force again soon.

Note: LD 50 for Boric Nitride is 2g/kg. Boron perked my ears, because we had 10 mg tube of different Boron based paste back in the 90s and it had potential lethality of killing whole engineering team. Luckily I casually mentioned that it seems to be highly toxic and our OSHA rep was in the ear distance. It was planned to be applied by fingers and glove usage standards were not the best back then. Apparently there was quite a back and worth and we eventually used the paste during prototyping (paste kept il a locked container, only trained personell using it and health advisory to anyone opening the device), but it was phased out to a more safe RF shielding paste.

@blackrifle6736

26 күн бұрын

*Oh, yes. Your story has the familiar defense electronics vibe, including the always-present OSHA rep. Mostly degree-flexing, annoying and intrusive, but priceless in this instance. Echoes of finding the ways to mill Beryllium MMW TWT cavities without killing the machinists in the process. Respect!*

glad to see you doing good , your kid is lovely , great video as always

I love how after Jay dropped the kryonaut he began using past tense to refer to it. "Welp that's under the desk now which we all know is the darkest of dark voids, yes it existed once but no more."

Most of the paste is supposed to be pushed out. You don't want any thickness. You want as much of the IHS and cooler plate to actually touch, with the paste only existing in the areas that will not touch because of microscopic deviations from absolutely flat surfaces.

@shanent5793

Ай бұрын

If you do that then you will get voids as the paste shrinks. The spec is usually one or two mils

@hmello3250

Ай бұрын

That is what I always thought but I had problems with Thermagic ZF-12 because of how thin it was. It works fine at first but then it squishes to the sides and temps start to get awful again.

@TheBigBlueMarble

Ай бұрын

@@hmello3250 there is a happy medium. If the paste is very thin then it runs out of the areas that it's supposed to be filling. Essentially leaving airspace in between the two surfaces. I guess I've just never run into one that's that thin.

@zadekeys2194

28 күн бұрын

​@@shanent5793 any proof of this? I've never heard of this, or any effects from this, if it happens. The guy you're responding to is right with some of what he is saying. You want to remove the air, using the thermal paste. So more is better (unless it's conductive). . Also, application is cold / room temp. From there it gets warm and it's viscosity increases, so the volume expants . Even when that contracts, I don't see it creating voids, it's not "slime" or playdoh.

@shanent5793

28 күн бұрын

@@zadekeys2194 I think it's common knowledge that thermal paste dries out and has to be replaced periodically. This is done because the oil component of the thermal paste evaporates, leaving behind grains of filler that have only point contact with each other and is thus a less efficient thermal conductor. So even if you managed to machine the surfaces of the heat spreader and cooler optically flat, and you put the perfect amount of paste so that all the points that can touch, do touch, and the paste only fills in the remaining voids-soon enough some of that oil will evaporate and it will no longer be the perfect amount. By putting more than the minimal amount it allows some reserve capacity between refreshes. There is a trade off between getting the lowest possible thermal resistance and a longer service life. Putting more paste increases the resistance but there will be more time until it has to be replaced. You can put less if you accept having to replace it more often. The specification of one or two mils (0.02-0.05 mm) is usually given by electronic component and thermal solution vendors

Man this was just a lot of fun to watch. Thanks for being you Jay!

These are the best videos: educative, informative and funny.

We need more testing on this. 🔥

I'm not supposed to let this out but that stuff is an additive to thermal paste and if you use it in a 1:5 ratio it will enhance any thing like kpx tgx etc. It beats anything by 5% fresh, you see about a 10% lift in cooling on older paste.

As you said about the conductivity, and its intended application, that TIM is probably best for applications that will regularly be maintaining temperatures well in excess of 100°C, unlike a CPU.

Jay explained it already in the passt but: W/mK=(W / m^2) / (K/m) Basically how much power per unit surface per Temperature gradient you get. Double the surface, double the power. Double the gradient, double the power. You can double the gradient by doubling the Temperature difference or halving the distance (usually the thickness).

Jay.. you and Steve are such AWESOME representatives for the tech community... We are blessed to have you (and your amazing crew) and Steve + his crew!! All you all rock always and in all y'all do.. All the time!! Praying for your health and the fam and crew as well!!

I personally used the Slice paste for my Slice Engineering Magnum+ that would regularly hit 300c for high flow rates and high temp materials; long story short it worked very well… interesting to see how it would work on PCs as that was something I was curious about. Thanks, Jay & the team!

Evaporate it a little before applying it to increase thiccness if that is a concern. However I think the reason its not ideal is that it isnt as smooth, most of the thermal transfer in the socket is metal on metal. The paste is there to full out the remaining gaps but if the past is grainy it might even reduce the metal on metal contact.

Could we get a part 2 where you mix the Boron Nitride Paste and Thermal Grizzly

jay & team, YOU GUYS ROOOOCK! glhf my friends!!!

As a pc builder and laptop repairer i would love to see the same sort of test done on a large range of the thermal pastes that are out there are the moment. My go to is the Arctic Silver and Noctua H1 but i have no comparisons as its always a new build or repair. The Arctic and H1 are "said" to be in the more premium range but not at the top price range. They both spread nicely unlike say Corsair XTM50 which is like a clay. Deepcool's that comes with AIO's spreads well but again no comparison. Maybe this could also include the new Thermal Grizzly KryoSheet. Hope your health is under control and keep up the good work guys.

So from what I know, you are kinda correct with 'thickness' (viscosity really), which is part of the deal here. That's why Noctua's paste has lower conductivity on paper but performs on par with other flagship products. There are other things to look at - conductivity might change with the temp the paste is at (some materials will conduct better when cold or better when hot). On top of that, as you mentioned - it's a paste, it has particles in it and the depending on how it behaves, it can have different conductivity in different directions (dimensions). And I'm just scratching the surface here on what's really going on...

Dude, imagine how cool this being your Bring Your Kid To Work day would be. You get to hang out with your parent AND play games all day. Hell yeah.

@duskshadow25

Ай бұрын

Yeah, no. If I want to game, I don’t want my parents around. As a parent you give them space and especially once they’re in their teenage years and going into adults too.

@RubyRoks

Ай бұрын

@@duskshadow25 Hard disagree. His youngest is like 9. Some supervision when gaming online at that age is a good thing. I say that as someone who was playing GTA Vice City unsupervised when it came out at age 5.

@Trocify225

Ай бұрын

@@duskshadow25well that sounds like you’re obviously doing or saying something you shouldn’t and don’t want your parents to know

@killerbsting1621

Ай бұрын

It's more like bring your kid to work week. Probably spring break for them. 😮

@duskshadow25

29 күн бұрын

@@RubyRoks I'm not talking about kids age. I'm talking about teenage, hence why I said "teenage years" in the original comment. If you have a child let's say 13 or under, then yeah, I would be paying a bit more attention on the games they're playing. But if they're in high school already, I would generally leave them along and try not to pry into their lives unless they approach you for advise specifically. When I went through high school as a teenager, I personally didn't like people budging into my own business. There's a thing call space and privacy, and it's one of those things I give people because I wouldn't want others to keep invading my own space and privacy. It has nothing to do about hiding something, but it has to do with letting your kids go and letting them grow and giving that room for them to decide what they want to do. And whatever they decide on, they need to understand there's a consequence to their actions. That's why I talk to them, instead of just pushing my own perspective onto them, as they get older. When your children are young, as in pre-teenage year, you can pay a bit more attention and deicide for them if they're struggling, but it's not like you can't communicate with them. I would still speak to them ask what they want to do, just like how you would communicate in your marriage. I personally don't think there's only one way to do something. Just like 1+1=2, 4-2=2 & 4/2=2 as well and so on. That's why you communicate. The the whole point of parenting isn't to force your perspective onto them but it's to give them the time and space to think for themselves. If they fall, don't pick them up (obviously talking about they can think for themselves already and not as a baby that can't even walk yet). Let themselves get up and figure it out. If they cry, that's fine. But it's gonna be okay and they'll move forward, because the sun will still rise and set tomorrow. Set them up for reality and success, not over shelter them with your own views. Because if you're overprotective of your kids and control everything they do, you're only going to push them away further. Something I can tell you is that, my parents were religious growing up. The praying never made sense to me growing up. But they gave me the freedom to choose what I wanted to do and decide my own path instead of them pushing their religious view onto me. I'm not a religious person like they are, but I still have good relationship with my parents because it's perfectly okay to disagree on something and still coexist. So many times I see parents force their views onto their kids and force them to believe in something just because that's way they believe or were raised to be. And people just blindly follow it and accept those beliefs without questioning or figuring out if it's actually right for themselves.

Been using GD900 from AliExpress for years...which costs $6 for a 30g syringe. Works great and costs pennies on the dollar relative to big name brands. I use it on both the CPU and the GPU, without any issues across the 4 PC's in my house. I do try to change out thermal paste once a year...as a matter of routine...not because temperatures started to rise around that timeframe.

The biggest predictor of a TIM's performance is minimum particle size, not thermal conductivity. Smaller particles, or fully liquid TIMs both allow for the filling of smaller imperfections but also reduce the distance between the IHS and the heatsink, which itself helps thermal transfer.

i tried diamond powder a mix, this also works better than expected.

This was a great video Props to the viewer Nick

I still would love to see a compilation of all the funny moments of the last years, plus outtakes.

Noctua doesn't share its thermal conductivity on their products because the measure can be affected by several variables, that could lead to unrealistic results.

Luv the look she shoots at you

Doing it the other way around and using PC grade thermal paste on your 3d printer can work (make sure it's good up to your printing temp plus a fair margin (10-20%) rather well. especially if you have issues with heat creep. Did that when changing out a nozzle once (from the same batch of .4's) because of clog. PLA and PETG all printed a smidge nicer on long prints after that.

This paste is usually used on hot end heaters. It normally doesnt have to hit 100°C Pressure isnt the issue as some hotends do have pressure. I use this stuff. It usually gets used around 70°C and i believe due to being water based it will dry out if kept too close to 100°C for too long.

Indeed that thermal loss between the heat spreader and the AIO. I have the H150i on my r7 7700x and its 31.7 on the coolant but 61 CPU using KryoSheet. However it holds at 93 on the ADA64 stress test down from 95 with the Corsair stock AIO paste.

This is too funny this video pops up when I just got thermal paste delivered.

@user-dv7hq2rh4g

Ай бұрын

For me personally it's either the cheap but almost as good as top end paste MX-4 or I go straight for liquid metal, like in case of GPUs or any other direct die cooling application.

@No-One.321

Ай бұрын

I literally ordered some thermal grizzly yesterday lol

@dmytrogalan1005

Ай бұрын

I mean, if we all survived until now without that gimmick of a paste .ight as well put our minds at ease and play - Buyers remorse and cope

@Alice_Fumo

29 күн бұрын

@@user-dv7hq2rh4g Gotta love me some MX-4. Can't beat that price/performance ratio

@diegoleiva7242

29 күн бұрын

Same, just bought some Corsair XTM70 which by the way is absolutely phenomenal for a scorcher of a laptop I've got. Brought the temps and fan ramping down quite noticeably.

My next build I'm going with the Grizzly Kryosheet and Grizzly Contact Frame

@brucepreston3927

Ай бұрын

I have been thinking about trying out the kryosheet myself...I'm using direct die with liquid metal right now, but it is very annoying to clean up...Plus I have to basicly lap my cold plate everytime because the liquid metal kinda etches the coldplate...I have heard the the kryosheet works almost as well as liquid metal and I won't have to worry about it anymore! I may throw one on my GPU as well...

@tmoore121

Ай бұрын

@@brucepreston3927 I don't think the Kryosheet works nearly as well as liquid metal in most benchmarks I've seen, but I don't think I've seen anyone trying it direct die either. That said, how often are you re-opening and re-applying LM? It should basically be a one and done job, it degrades maybe a couple C over years.

@TTM1895

Ай бұрын

I've never used liquid metal, so I can't speak to whether or not it is better than the Kryosheet, but I have been using the Kryosheet since it has come out, and it works very well.

@brucepreston3927

Ай бұрын

@@tmoore121 My PC gets very dirty where I live so I usually do a full teardown and deep clean twice a year...I know I don't need to change the LM every time, but I do it just for piece of mind and because I might as well if i'm already doing a full teardown of everything else...I will pick up a couple kryosheets just to try it and see what temps I get...I know it won't be as good as LM, but if it's pretty close I may just keep using it...I'm sure it will work better on my GPU also, I haven't been brave enough to try LM on that...

@brucepreston3927

Ай бұрын

@@TTM1895 Thats good to hear! I know it won't be quite as good as LM, but it will fun just to test it and see what I get...If it works well in my PC, I may try it on my laptop also...I just like tearing computers apart! lol

Once I used anti-seize as I was out of arctic silver. It actually worked quite well!

@soaphelps

29 күн бұрын

the pure silver stuff works very well.

Well, that was cool. Learned some stuff. ✌️

Honestly these mad scientist sort of experiment videos are the best! Even if it fails, still a good ride

Oh hell yeah 3:09 love the overlay!

simple answer is its to jay term - LIQUIDYEE for use, which is why its not used ever in such applications as cpu cooling, This is why water based thermal stuff is never used on those parts, its used where u can place it in a hole or area that wont be presurized to make it move once it dries due to its viscosity. it needs to dry solid, if pressed it wont be thick enough to work, it will get pushed out as saw in the video.

I use Prolima PK-3 for thermal paste works really well.

Der 8auer explained this in a different video pretty well: thermal conductivity is highly dependent on the temperature difference and temperature overall. High temperature mean a higher conductivity, no matter what. And manufacturers measure it very differently and are free to choose whatever method suits them best and produces the highest numbers. Thermally stable materials will then be tested at the highest temperature they can sustain, because that produces a higher number and better marketing material. A paste for CPUs is often tested at 100 degrees centigrade which is the CPU maximum, or sometimes at the highest temperature the paste can do, 200 or something, because the conductivity will appear to be higher, despite the CPU never reaching that temp and hence, conductivity, ever. A paste meant for 3d print heads will be subject to 200-300 degrees centigrade, and thus the conductivity is measured at that temperature. Which a CPU will never reach, and the effective conductivity for CPU application is therefore much, much lower. Unless the manufacturer specifically states at what temperature their conductivity was measured, assume the absolute worst of them. Assume the manufacturer tested at the maximum paste temperature or 100 degrees, whichever is higher and the actual real-world conductivity at your lower real-world degrees is MUCH lower than that. And there is no best paste for all systems, because desktop CPUs will be running between 35-50 idle and laptop CPUs to 50-65. Desktop CPU users will call the fire brigade and the emergency shutdown when the desktop CPU reaches 99 degrees, while 99 degrees for a gaming laptop CPU is just a regular Tuesday.

3:18 Thanks for clarifying the temp in the room isn’t 70° Celsius.

Great video as always, hope you are doing well. Could you do these tests also on X3D chips from AMD, so we can see how it does with those cpu's. Thanks for your hard work on these video's.

FYI, boron nitride is also used as a non-wetting agent in aluminum foundries.

Jay, VID doesn't droop. VID is the voltage the cpu is requesting from the VID V/F curve table. VID isn't the per core voltage. Each core and ring is requesting a voltage for their frequency at that moment. The VID sensor is reporting the highest voltage being requested out of all the cores/ring. With boards with die sense, the vcore sensor is the actual core voltage after vdroop. With boards without die sense there its best to use VR VOUT.

I've noticed one thing about expensive thermal pastes. The best ones spread super easily and that's what you should be paying for. You can get a big tube of PK-3 or something pretty cheap, but something like a Kryonaut that's a couple times more expensive... way easier application. Good application = good performance. You can get there easily or you can find a spatula for the thick stuff.

@SolaufeinG

26 күн бұрын

very good. someone got the point. the application is more important than the material itself. conductivity is not "linear". the 13.2 vs 12.5 conductivity won't have any effect. Even beyond 20 W/mK is not something spectacular. This is a UNIT of measurement. Not a guide. Watts per meter Kelvin. written wrong from the beginning ... Joule was like "am i a joke to you?"

A long time ago Vince did perform essentially a viscosity test. They called it film strength back then, but as I recall the other dude involved did state that a specific range of film strength does make a difference. While they were chasing literally 2 or 3 degrees... It still made a difference! Great video dudes! And hi to yer CEO, I mean daughter! Lol!

Little Jay the boss in the background 😂 making sure you're reading the script right?

FYI for anyone out there that thinks you can "overtighten your cooler and squeeze too much paste out" that's not how it works 99% of the time. any paste that can be squeezed out should be. direct IHS to cooler is the best cooling, paste is just there to fill the microscopic spots where they can't make direct contact because at the end of the day we can only make things so flat

Please do a series of videos or a big roundup video of a bunch of different thermal paste types! But you know, Jay it up the way you do lol

@michaelbraaten

19 күн бұрын

And I mean actual thermal pastes, not things that aren’t, like peanut butter haha

Try mixing that Boron nitride to Aluminum nitride + a bit of mx-2 and see if it will enhance the structure of the thermal paste.

Talking about things being really high. I went to crawl under my table to grab the grizzly paste as you dropped it

11:52 *You can just fill that whole void with the white stuff* Beautiful. Paul’s should be proud.

Are you down with the thickness?!?! OOOOOOOO-AH-HA-AH-HA-AH-HA-AH

The thermal paste that I use is called "HY-510". I bought 300g of it in 10 30g plunger tubes for all of $20CAD. How well does it work? My Ryzen 7 5800X3D is cooled by an AMD Wraith Spire which I installed about 1¾ years ago using HY-510. So, here we are 1¾ years later and, according to Speedfan's reading at this very moment, it's holding nicely at 42°C. Not bad for a CPU with a max temperature of 90°C. It really doesn't matter what thermal paste you use unless you're some crazy overclocker because for typical gaming and home use, my HY-510 works just fine. I have zero complaints.

It would be interesting to see you test the new thermal pads LTT is selling and using in their videos.

So one thing that's very important for thermal paste is a very low surface tension and viscosity because you want to make the layer as thin as possible. I'm guessing that the boron nitrate is not nearly as viscous as the thermal paste, so it doesn't allow for an incredibly thin layer that forms to the both surfaces perfectly.

I just started using ptm7950 on everything since it's easy, lasts like forever, doesn't pump out, and get close to Liquid Metal

So the question now becomes "would it have worked better if it was more viscous? And how would that effect the thermal conductivity? And electrical conductivity?" We should get derbauer on it.

@barongerhardt

Ай бұрын

IIRC, there is an interview with Der8auer where he calls the thermal conductivity ratings are junk. That they can engineer and game the number to what ever they want and it will have no noticeable benefit in real world testing.

@MrLunithy

Ай бұрын

Engine oil does a great job in a car as one of the three jobs it does, plenty of viscosity there.

One alternative that actually works is Never Seize from the auto parts store. Used that for a whole summer of builds back in the Pentium 4 days.

I would be interested in seeing a comparison between the Honeywell 7950 ptm vs the thermal grizzly thermal pad. I think it would make a very interesting video.

I use Prolimatech PK-3 Nano Aluminium, it’s 11.2 W/mK and it works a dream on both my 13700K with Noctua NH D15 and also my HP Omen laptop, 10750H and 2070 MaxQ.

I have used a few different types of paste over the years, but the last few RYZEN builds I have used thermal grizzly Aeronaut, my latest build has a RYZEN 9 5900x with an AIO.

That reminds me of the type of paste used for mosfets/transistors heatsinks, it's not normally compressed like a cpu heatsink

I like that we have some back ground gaming in this video

I wonder if a thermal pad could be cutout to act as a bib around the CPU heat spreader, to make ease of cleaning old paste more simple AND to avoid any leakage of excessive paste onto the surrounding CPU PCB elements or socket perimeter. A bib that doesn't interfere with the cooler contact area.

IHS with water cooler interface build in or lapping the IHS and cooler together, basically rubbing it against each other with some polish paste in between should be the best solution regardless of the thermal paste. It would be maybe even possible if lapped correctly to use a very thin graphene layer not the thick stacks.

Id like to see this used on some of the new ThermalRight ,DeepCool and Arctic coolers, which seem to have very good quality contact pads to the ihs.

I think part of the problem is that the boron nitride paste is really designed for use at higher temps than the pc will ever reach. It's designed to be effective in the 200 C - 400 C range that 3d printer hotends operate at. I think it also is supposed to have time to dry out before it works at it's peak efficiency, which probably would have required a few hours or maybe days.

it's fun and games, and easy content to compare different TIMs right after the application, but not indicative of any real-world use case. I'd love to see a test over a longer period of time. like a year or three, eventually 7. Find out what works in the long run, not just for a nice benchmark score.

Try Gelid GC Extreme. Fantastic in general, but especially for GPUs. Recommended to warm it up before trying to apply it.

The suffering with this stuff Jay, is that its a water-based solution; when it dries and the water evaporates it turns into a sort-of-weak-cement of sorts. Most people using it on 3d printers tend to bring it up to a boiling temperature slowly to evaporate the water and avoid bubbles in it/etc.. Re: YOU NEED TO LET IT DRY SLOWLY TO WORK PROPERLY.

The air will always form pockets within the paste, albeit tiny ones, because unless the paste is permeable to air, the air remains trapped in the tiny imperfections of the paste and the two surfaces. Heating and cooling the system will pump out the air, probably, and this is PROBABLY what we experience as "burn-in" period of newly applied paste. Still, paste will not have lateral heat transfer nearly as good as solid material. That's why integrated heat spreaders remove hotspots and get the CPU die to a more homogeneous temperature, which then is lowered by the copper and the fans. Direct-on-die application of thermal paste cannot easily get rid of hotspots unless it's liquid metal TIM. For notebooks, liquid metal is a suicide pact if something goes wrong or the notebook is moved when hot. Which it will in a few years when you have forgotten about it for one minute. Laptops thus can't be cooled with paste of LM TIM well enough. After trying it on 3 different ThinkPads, I can recommend the Graphene pads, which perform similarly to paste in bulk heat transfer, but totally exceed over paste in removing hotspots. The maximum wattage for sustained benchmarks and loads is the same or better than good paste, but within the 1-3 degrees centigrade measuring error. However, single core loads don't spike rapidly and don't trigger the throttling. And application is done in a few seconds, unless one is a perfectionist. Then you'd lap and polish the copper plate to a mirror shine (30min, up to 10,000 grit) and degrease everything with iso alcohol and then lay down the graphene pad with soft plastic tweezers.) never touch them with bare hands, though.

Locked the voltage but left the LLC :D Also there are at least other 4-5 power saving options in the BIOS that you have to remove if you want to have somewhat accurate results. Not just... in the ballpark.. I see the frequency is not asctually locked as well

Could you make a video about making your own paste with boron nitride powder and silicone oil?

I'm just thinking what about ceramic paste or copper paste used for brakes to use for CPU coolers? And about the Boron nitride paste - as far as I remember in the manual for my 3D printer was written the paste has to dry before you are allowed to use the printer or the thermal conductivity doesn't work.

I sometimes wonder if Little Jay has ever rewatched the drunk livestreams Jay has done with Barnacules.😂

Well, you don't have to go into to what W/(m⋅K) means, as that is what it means. As long as the thermal paste has decent thermal conductivity, small differences in the thermal conductivity between different pastes matters little compared if there's air pockets, and whether or not the layer of paste is unnecessarily thick. There should be as little paste as possible, while making sure all voids are filled, as thermal resistivity increases with the thickness of the paste layer, no matter the W/(m⋅K) value of the paste. Thermal paste is truly one of the cases where "less is more", as long as all the air is replaced by paste. A 2 micron thick layer will have twice as much thermal resistivity as a 1 micron thick layer, all else equal.

Little Jay in the office, make her earn hose PC upgrades. Make her the new head of marketing

@GrySgtBubba

Ай бұрын

"Hose PC upgrades"...? Water cooling unit? 😂

@ronaldlebeck9577

Ай бұрын

Maybe testing would be better...trying to make something kid-proof would be nearly has difficult as making something sailor-proof (I'm a Navy vet and I know for sure that *nothing* is "sailor-proof"!) 🤣

I like to wait on boron nitride to cure first, I like to wait about 24 hour period when applying it to my heat break on my E3D v6 hot end.

So we finally found a reliable source for the Honeywell phase change thermal pads, which are great. We've got carbon thermal sheets which are also great. And now we get boron nitride super paste which I assume is also going to be great What a time to be in the DIY computer space How long before liquid metal just becomes completely unnecessary? To the best of my knowledge, it still has the best thermal conductivity but for how much longer? Upon further, reading into the boron nitride stuff. I don't think I would ever actually recommend using it in a PC just because of the preparation requirements, but it's interesting and I'm going to keep using it in my 3D printer but It seems to be less usable than liquid metal for less benefit. And it is significantly less convenient than the graphite and Honeywell thermal pads

That easy to heat it to 100c by heatgunning the heatsink and let it cool it down for awhile and test it

make sure u clean that off once that stuff is set it it pretty will act as super glue and almost will "weld" anything together

Mix that paste with some of the KPX and see what it does? The KPX being a more solid medium might hold it together better...

I had Arctic Silver on my old 4770k for 8yrs before i decided to redo it for fun, Was rebuilding soon after anyway, Only 4 or 5c difference on Intel burn test or Cinebench. 86c down to 81c max, Faired pretty well for all that time, 7700x now with Grizzly.

Can you try to mix it together with the KPX to make it thicker and see if it changes anything?

Wonder if the Boron Nitride just overwhelmed the cooler... maybe would work fine with CO2 or LN cooling?

Even when truthful thermal conductivity numbers can’t really be directly compared, you can easily “game” the numbers through testing methodology. I know Roman has talked about this as have others. So big number not necessarily better. Also, seems like the limiting factor for a lot of CPUs (especially Intel) is the heat density and getting the heat from the die to the cooler. So much power density and then the IHS..

2 things conductive when wet and can leave Corrosion .. and it will dry out , and you can always just heat up the block apply some paste on block and pre heat

For comparison, I just looked up that the thermal conductivity of air is about 0.025 w/(m-k). So even the Thermal Grizzly is already 500x more conductive.

The chemists might know more about thermodynamics but they fall short on the verbal hilarity. 😅 Loved this video Jay.

I love the humor in your vids so much. The spontanous / situational comedy moments are the gold next to the interesting information and insights.

Is anyone else dying for a new J2C pc build vlog? 😊