Це відео не доступне.
Перепрошуємо.
5 Watercooling Myths You Shouldn't Believe
Вставка
- Опубліковано 5 вер 2024
- These are just some of the 5 most common myths I hear on a regular basis regarding watercooling your PC. Think I left something out, then be sure to let me know in the comments section or twitter!
Visit www.coolermaste... to learn more about the MasterKeys Pro RGB Keyboards
○○○ All music provided with permission by audiomicro.com ○○○
SUBSCRIBE! bit.ly/sub2Jayz...
WATCH MY MOST POPULAR VIDEOS!: (link)
••• Follow me on your favorite Social Media! •••
Facebook: / jayztwocents
Twitter: / jayztwocents
Instagram: / jayztwocents
About JayzTwoCents
Your one stop place to shop for fun and educational videos on all things tech! Oh, and don't forget gaming! Gotta make time for gaming!
5 Watercooling Myths You Shouldn't Believe | JayzTwoCents
• 5 Watercooling Myths Y...
JAYZTWOCENTS
/ jayztwocents
Do mythbusting on overclocking
DO it
damn that hair on my screen LOL :D :D :D
i dare you
YESSSSS
+1 for this :-)
1:44
I like how you're sitting on the stairs. It's like you're talking directly at me. So casual. So cool. See everybody? He is in his 30s, but he understands what the kids can relate to. I like this. I like you.
Lol?
k
Nani?
I don't give a shit about stairs.
Give me a Gaming Chairs. LOL
the fuck are you on about?
Jay could you cover myths about overclocking gpu and cpu next?
I third this!
Why? He has no idea what he's talking about.
@@spudhead169 are you sure
@@saab35draken39 Just have to watch his videos. Like the one on water cooling "myths" where literally everything he said was wrong.
@@spudhead169 *_ O H _*
5 myths about myths plz
1 - Myths are usually not true.
2 - Myths are close relatives to a lamp worshipping species of insects.
3 - Myths tend to be busted by a pair of funnymen.
4 - Myths start out as simple misinformation and fundamentalize over time.
5 - Number two is the utter truth and definitely not a myth.
Alternative #2: a myth is a female moth
Myths is not facts
damn and here I was thinking that the algae growing in my system was helping to promote cooling
Forbesbm124 green colour for free
Forbesbm124 maybe you should put some snails or fish in there that eat algae ;p
That's the out of the box thinking I like to see.
maybe you could feed the algae with co2 from the atmosphere and release the produced oxygen to improve your systems ecological footprint
Forbesbm124 s
One myth I've heard that I've actually been recommended to do is flashing a motherboards BIOS twice with the same version. For years I've heard it was a bad thing & easy way to brick it but on the MSI Forums it seems a common practice to fix bugs.
"...bigger Reservoirs don´t increase your cooling capacity..."
Me looking on my pool-sized radiatorless Reservoir: "Dont cry my little Water, Uncle Jay just make a bad Joke."
yep he really doesn't know what he's doing. Just watch his skunkworks build. Putting in dead motherboard, wrong pcie slots for x16/x16, using AF fans on radiator, blowing hot air back into his case etc...
@@donjones2500 That was obviously a joke for hyperbole, it doesn't increase your cooling capacity per se, it just makes it takes longer to equalize. Youre the idiot here taking a joke about a pool serious lol. When you have a million motherboards and one is dead, can you really be so nit picky as to blame someone for picking the wrong one? You really want to be angry huh? He's talked numerous times before on what kind of fans for radiators and that you want static pressure blah blah (in videos that are, ya know, actually supposed to be instructive), skunk works is old as hell, you should prolly just go back to your grumpy hole..
Meanwhile my stock cooler cries in the corner. =(
Don Jones I’m betting you have a huge reservoir not knowing how thermal dynamics with water works
It also isn't "radiatorless" since the surface area around the swimming pool (walls/floor/surface) is the radiator of that system and has a huuuge surface area and the thermal conduction to the ground through the walls of a pool is really high
once you see jay's face in 4k you can never go back
@S4mandar the human eye can only see up to 4 gb ram
4k is so dumb homosapiens eye can't see more than 144p smh
@@mango251 Just download what youre missing
they human eye can only hear 1 gigawatt of 3 gb of cpu vram
Top 10 uwu moments
I haven’t subscribed yet but with videos like this where you’re testing and showing REAL WORLD results. I can’t help but want to be a part of your space. Thank you, Jay.
Loved the change of scenery between myths.
Looking forward to your loop order video. Love seeing your builds. They are just wonderful. Keep up the great work.
id like to hear about parallel vs series loops
Can you do a cost/performance analysis for water cooling? Its been forever since I even looked at parts so I just don't remember how much they cost. Could you tally up your system and show just how much money is spent on the overclock, then compare it with an air cooler? Is the potentially hundreds of dollars worth it for the extra few megahertz on an overclock?
Considering high end air coolers still cost less than aio water coolers, and have equal or greater cooling performance, water cooling holds no value in comparison. Air cooling will always offer the most value, followed by AIO water coolers, then custom water cooling loops, then hardline loops are way below that. Hardline cooling is A: expensive just due to the fittings alone, and even if you go with more bends vs. fittings to minimize costs, the added time it takes to make the bends, or to do them all over again when you make a mistake greatly increases the cost as well, depending of course on how much you value you're own time. And B: the amount of extra cooling it provides doesn't always give good results, just look at Ryzen and Pascal GPUs, they don't get that hot to begin with so power delivery and it's own limitations becomes the bottleneck.
So basically its just expensive case bling like cold cathode and random LEDs everywhere? I mean I understand water cooling provides a bit more stability in temps but its hard to see why anyone would go water cooling, even AIO coolers when similarly (and cheaper) air coolers can handle everything except harder overclocks.
I believe he's done that video once before. It was way back though maybe 2 years ago. EK's new aluminum custom loop set up is a good deal and new to the market, I think Jay has also done a video on it as well.
It's all about lower noise levels and aesthetics these days. Ten to fifteen years ago, when watercooling was a DIY hack and 120mm case/heatsink fans were rare, it made a much larger practical difference.
Sensibly done custom loops can actually be done at a reasonable price and are more about reducing overall noise by oversizing the radiator and undersizing the fans. Think reaching sub-ambient noise deltas. Performance wise there's little difference between water and air and there never was much. Yeah. For a VERY short time, there were no big air heatsinks or decent air coolers on GPU's... It didn't take the AIB's long to figure out how to design good coolers, or for CPU cooler manufacturers to start to grow out of former DIY'ers getting inundated with requests, and prior existing ones jumping on the trend after the fact.
But yes. There are certainly bling builders who do blingy water loops. Jay is one. He's more about bling than function, actually. I mean... Look at the goof up he did recently with the Ryzen build by going for bling and actually HURTING his performance.
0:40
" *hey Charles, can you press this button randomly for me?* "
" *certainly, sir* "
The reason WC systems may heat a room up more is you can push more power through the components without hitting a thermal throttle. As far as heating your room, our PC's are essentially low/medium powered space heaters (depending on build). The wattage they draw from the wall is released into the room as heat. If an air and water cooled system draw the same power from the wall, they will heat the room the same.
or because the average water cooling radiator, lets say a 240, has more surface area to dissipate heat than an average air cooler.. the purpose of a radiator/heat sink is to move heat away from the cooled component. away means into the ambient air. more heat dissipation = lower cooled component temp and higher ambient temp. a fine example would be something like a noctua d15, which has pretty similar total cooler surface area to a 240mm AIO and can maintain about the same temps.
You're on the right track. The first law of thermodynamics tells us that energy has to be conserved. That means all the power your system draws from the wall is being put into the room, regardless of the cooler type. As long as the system is consuming the same amount of power, the room temp increase will be identical regardless if using an air or liquid cooler. For example, if a d15 and h100 are both able to keep the chip cool so it doesn't throttle, the increase to ambient will be the same.
The reason I said liquid cooling might heat a room more is due to the fact it is likely run higher clocks/voltage/power consumption since it won't thermal throttle vs a stock air cooler. Since the system would be drawing more power from the wall, running the higher voltages and clocks, the ambient temperature would increase.
This is not how thermodynamics works. Heat in, heat out. Doesn't matter how "hot" your cpu gets; if its putting 150w of heat into the room, its 150w regardless of what cooling you have.
a larger reservoir doesn't affect max temp, but it does slow changes in temp, reducing swings in temp that cause mechanical stress.
What if the reservoir was made of copper. Like a 3" diameter copper pipe cut to 8'' length and reduced on both sides for fittings. You would get some heat transfer with air movement in the case. Correct?
@@deesmods6696 With a large enough reservoir, the reservoir itself could function as a radiator. So, while in most cases he's probably correct here, in an absolute case, he's wrong!
Love these type of videos. Keep it up. Proud of you
h3h3?
dennis045 keep it up!
YO WADDUP HILA KLEINERS
I liked the moving around of scenes, especially the stairs. Feels candid and honest. Not that you don't usually feel honest....but yeah.
Jayz2cents spittin that watercoolin game in back alleys and shit.
Man I can't believe how much better jays videos look compared to the other content creators I watch. Thank you jay !
next, can you go over Overclocking your GPU. the other one is kinda outof date
it's not out of date at all still applies
I'm lovin these myth vids...You bare the burden of having great knowledge coupled with the ability to impart it simply to the masses; the masses are pleased. Can't wait for the bits showing the varying loop designs and the results. Thanks, Jay!
how old is jay his hair makes him looks like hes 50 but his face makes him look like hes 25 and he acts like a 10 year old
42
yoda won about 40ish but he could also be dyeing his hair, I've seen tons of people using grey/silver hair dye recently
you won this battle Jay but you really don't seem to be 42 if you're 42 i thought you was like early 30's to mid 30's. and i always see you with grey hair if that's your real hair color. and you still like to act like a little kid and be serious at times to so i was always thinking you was around that age group. that's why i like you jay cause of your kiddie side and your love of computers and cars
jay did a draw my life a few years ago and gave us his birthday in that video. look it up. interesring story too.
He's 36, if I remember right. He just had a bday not too long ago. I thought he was older me. Nope. He's younger.
Unfortunately, when it comes to large overclocks, loop order does matter. I had a build where a 7700k would go unstable if the GPU was in the loop before it as that added temperature did take the CPU just a few degrees higher and lost stability, Switched them around and it was rock solid.
I'd like to see a video where we can learn more about custom loops and tube bending. It's really intimidating as a beginner for water cooling but looks amazing.
Hey Jay. Great video! I just want to say I feel like during the radiator surface area myth you skipped over a good chance to cover, and perhaps may have even confused some on having enough surface area to efficiently cool. You were completely right on it cooling better beyond a certain point. What I'm talking about is having too little surface area. In turn we're actually talking about inefficient cooling and not impossible efficiency, but, some may not know the difference.
Please this "myths" a series. :)
I am a Mechanic, and I can tell you that the size of the reservoir is ultimately crucial to the cooling capacity to any liquid cooling system. Whether you are dealing with a CPU/ GPU , or a nuclear reactor, the amount of coolant is the key. Water tends to hold heat very well (Even better than Iron). Also, speeding up the pump might not be the best idea, as the liquid in the radiator needs time to cool down. Air is a very poor conductor of heat.
That being said, listen to Jay, (and those like him that are trustworthy), cause they have already done the research, and made most of the mistakes that you should avoid. (Be it for cost, pride, or both) I'm no different.
Just remember that "Everyday's A School Day !"
Thanks Jay!
Matt
Jay please do a tutorial on how to change out the fluid in your loop
Floating Paradise I think it depends on how the loop was set up. If you set up your own loop you shouldn't need someone to tell you how to empty it. If you had someone do it for you, that should have been the first question asked. Just my 2 cents.
Grey Roots Not necessarily. I always have trouble draining my loops. I can never get all of the fluid out before having to disassemble. These videos make it look so easy..
Scott G. Helps to have a drain valve at the lowest point.
SkinUpMonkey I know. I always have one, but water is still always in the blocks, or doesn't make it around the bends in the tubing.
Scott G. For blocks make it so the block is flat then tip it around about a bit. Watercooling is all about the tipping lol.
Not for nothing but these videos really do help me. I've always wanted to dive into custom and I'll be honest I've been afraid and have always settled for aio coolers but for my next build I'm thinking of buying the aluminum ek that you reviewed and giving it a shot. Thanks Jay.
For the water to heat up say your cpu after having gone through your gpu block the water would have to be at a higher temperature then the cpu. So it's technically possible but if thats a problem that you have then I think you have other things to worry about.
Not necessarily. The bigger the difference in temperature between two surfaces, the faster heat will transfer from one to the other. Meaning if your water temp is 50° below your CPU temp, it will conduct away a whole heap more heat energy than if it were say, 20° below your CPU temp.
This is way better than the other one. For me you're my number one source of UA-camrs in the water cooling work. You should do more videos like this. Thank you man!
@JayzTwoCents can you test a water-cooled parallel against a series loop in regards to cpu & Gpu's temperatures, water flow rate and If it has any impact onto noise generation?
Well, thermodynamics state that heat in vs heat out is a net function; the laws of physics dont care about parallel or series loops. The ONLY consideration is head loss, but it winds up being a very negligible.
+JgHaverty I think you misunterstood that law. Heat is not temperature. Heat is energy. 1kg of 20°C water has way more heat energy than 1kg of 20°C air. You can only equalize temperatures and not the heat energy.
If you have a higher temperature difference, you can equalize temperatures faster. In a serial loop, you pre heat the water for the next block and in the radiator you pre cool it down for the next radiator. So the difference become lower and you get higher temperatures.
So the energy that goes in and the energy that stays in the water and leaves it, are equal. But with a serial loop more energy stays in the water and it becomes more warm.
I hope i explained it well, english is not my main language!
This is all above my pay grade...
I like that Switching of spots. Especially that spot on the stairs goes very well with that kind of Vidoe.
¿Is it true that even very good quality thermal grease dries out in time, and you have to change it every 6 months or so?
No it is not true. If you want to know IF you need to change the thermal grease.. Watch your temperatures.. If they start to go up over time that would be one of the things you check. If your water cooled you check other things like in a custom loop - is there something wrong in the fluid? clogs in the water block fins?
There is no need to change thermal pastes or anything else if your temperatures are where they should be + - changes to your room temperature..
Take this example: You put your cooler on in winter.. Your room is like 68f or 20c.. 6 months later you see your temperatures going up.. It is summer and your room is now 78f or 25.5c.. So your CPU/GPU temperatures will now be higher because your ambient temperature has risen.. This is not a fault of the Thermal Paste.
Replace the thermal paste every time you remove the heat sink.. It is more likely to have temp issues from dust and crap that builds up then from thermal paste degradation.
Now if you are removing your heatsink / Radiator every 6 months to clean it, then yea replace the thermal paste at the same time.
TisJester XxX thanks man. That is actually very helpfull. I Have a 212 Evo with a 6700k without oc. My temps are great, but I had that doubt. I dont remove my cooler often because i have little to no dust in my PC, and just clean The fan when I need, without removing the heatsink...
Noctua said their thermal paste has life on shelf for 3 years and upto 5 years on your CPU
Not sure about the expensive stuff, but my standard paste is dried out when I change it about every year and a half or so.
If had an intel i7 920 over clocked to 3.3ghz for 10 years in constant use I've only replaced it once on a stock cooler have never had one problem.
Having water cooling that keeps your rig at a stable 40C while gaming and around room temperature when not gaming is going to keep the room cooler than if you're running a cooler that keeps your system at 60C while gaming and 40 while idle.
I've actually experienced this, as my gaming room got a few degrees colder after longer gaming sessions with the water cooler.
Eventually both systems would heat the room up to the same temperature, but it would take much longer with the watercooling and people don't generally game for more than 8 hours at a time.
A large reservoir also looks better
That's an opinion and not even always true. A large reservoir in a tiny case looks cramped and not necessarily better.
Wade Bingham some of us have big cases so need to fill the space with something
exp10der *A large reservoir looks better except when it doesn't.
I prefer one radiator per CPU or per GPU so that I can swap cards. I like the reservoir integrated into the radiator solution to save space. However, no manufacturer makes such a system for custom make.
i see non-conductive fluid being best for people who have never done watercooling before so if they messed up a fitting or didnt connect a tube good enough its not gonna ruin anything when it leaks while they are filling the loop for the first time or the first week of the loops use.
A video about flowrate would be interesting.
I agree. You could probably find the exact amount of radiators that would guarantee the most heat removal given a specified pump head.
We get it, you vape.
Flow rate won't really affect your cooling much at all, unless you literally turn the pump off. One could argue that a lower flow rate would cause the system to take a bit longer to ramp up to a stable temperature (like if you have a larger reservoir), but I think the difference would be negligible. In terms of max temps for your components, you wouldn't see a difference, because even with a low flow rate, all of the water in your loop will be roughly the same temperature everywhere (within a few degrees at least). That temperature that the cooling liquid stabilizes at will depend on the heat output of your components, the ambient temperature, and your radiator.
+Sebastian George, The argument about flow rate isn't about how long a given bit of water is over the heating/cooling components but rather how that water moves over that component. It's a fluid dynamics thing.
If the water is smoothly moving over a flat plate and most of the water molecules are moving in a straight line, the effective surface area is much lower than if the water is turbulent over that plate. The more effective surface area you have the more the gpu/cpu is cooled.
Much of this is handled with well-designed water blocks, but each water block will have some specific flow rate for optimal heat transfer. Now how important this is in a modern loop, I don't know, but in theory it could be very significant.
It's kind of like if you used a thicker block (more metal between the cpu and the water). It would take longer to remove the heat and, despite the loop being likely the same temp, the cpu would stabilize at a higher temp. Same with poor thermal paste.
+slicedtoad I agree that it COULD be a significant effect in theory. However, I don't think it is in reality. If you open up a modern water block, you'll see that directly about the CPU/GPU, the water is forced through a bunch of very narrow channels cut into the metal. Given the viscosity of water and the sorts of flow rates that you'll get from any PC watercooling pump, the flow through those channels will be pretty much laminar (i.e., not very turbulent at all).
Not only that, but the water flows past the CPU/GPU fast enough that its temperature is only raised very slightly, regardless of whether there's turbulence or not. You constantly have new water molecules coming into contact with the heat sink, but it's because you have a net flow, not because of turbulence. Like I said above, it might make SOME difference if you have a super low flow rate, but I don't think modern watercooling pumps can give low enough flow rates to cause a significant effect without being turned off.
This video has helped me out so many years later. Thank you
Surely if I stick the radiator into my freezer it would cool below ambient room temp??? 😂
Jas George for a short while, probably. If you live somewhere cold the trick is to stick the rad outside of your building 👍
Jas George bitwit did it with an ice water bath for the radiator and it did work. Linus already showed that putting the heat into a refrigerator or freezer won't work as it can't handle the heat being put out by the components.
It would destroy your fridge at the same time. Linus tried that one for science and gave a mini fridge a major hernia.
Depends on the output of the compressor, the size of the condenser and radiator in your fridge. Realistically a consumer fridge or freezer will not have the capacity to do so as most of the time they have less heat transfer than a small A/c, proof of which is seen when you hook up your new freezer and it takes forever to freeze stuff.
just use a peltier cooler on it. sub ambient for cheap
I have been reviewing many tech youtube channels and after looking at over 30 channels this guy is the best. My next favorite is the Science Studio.
I am thinking of buying a proper pc case from caselabs. A Magnum TH10A or SMA8 would suffice, but i am thinking #yolo because TX10 is back for a limited time. Might aswell pay the extra few bucks. I had a watercooled PC 6-7years ago and then i used a old Thermaltake Mozart TX case which i did some light modding on. I bet any caselab case would be great for waterbuild. But i am a bit facinated by TX10. Have 1.5weeks to decide whether to buy it or not.
Whoa @ the TX-10. If I had that money right at the moment, or was likely to have it before they took it away again, I'd totally pick one up. A -D version. Possibility to fit up to three systems in one? Connect all of them to a single epic oversized radiator loop? Hells yeah.
My only gripe about it is that they don't seem to have as an option a desktop style case that can fit HPTX boards. In fact, the TX-10 is the only case they have that can fit HPTX boards at all, and they don't even sell it consistently. What's up with that anyway? Why would you do limited runs only of a mega case as your only case that can handle larger systems at all? A somewhat more reasonable case could easily be designed for HPTX, it just takes a few extra cm of depth. Why even limit availability of the mega case? I understand only doing limited runs of it, sure, but keep it permanently listed, and when you run out of stock, do another limited run. It's not rocket science to keep availability of cases like this. They aren't going to sell by the hundreds or thousands, of course not, but they'll sell, just more slowly.
What are you putting in it? I'm a big fan of finding the smallest possible space to accommodate the components.
rosewell1983
Why? Their cases are no nicer than a Fractal and are 5x the cost...there is ZERO reason for anyone to buy one of those.....ive built in 6 of them and all but 1 of them the customer switched out to other brands..... The other one didn't because he got the case for free Pre- modded from a giveaway
Really? I've never heard of, nor, when I just tried browsing Fractal's catalog just now could I find, a single case from Fractal that could support anything like the modularity of a CaseLabs case. None of them can handle a repurposed automotive radiator without extensive re-machining of the case itself. None of them can house more than one system in a single case. None of them allow swap ins and swap outs to the extent that CaseLabs demonstrates. And they're a helluva lot more expensive than other cases equally as dismal at modularity.
Thanks Jay! Going to do custom water loops for the first time on my next build and this info is great, looking forward to the follow up vids as well!
OMG jay, can you do a loop with a refrigerated rad????
like with one of these
thereefshop.com.au/index.php?main_page=product_info&cPath=43&products_id=756&zenid=5abf43e21a7a321b82649dbc75546fee
Robert Goodfellow you can but you would need a wider pipe or a converter for the connections to a fridges rad you would also need to cool it
adamania its not a fridges rad its a cooler for an aquarium and there are ones specifically for computers there are already a few youtube videos on them (go search it lol) but the videos are poor quality, no where near jays level, and they also dont try overclock or show any benchmark results.
Funny video. Now we need a myth busting video to clarify the new myths introduced in this video.
My favorite was: water cooling increases your room temperature more because it is more efficient at getting the best of of your pc.
Ymmd!
what about pump speed myth?
I was typing a comment about doing a parallel loop as you said it, please make a video explaining the benefits/draw backs of parallel vs serial loops!
I am pretty sure watercooling isn't going to make your room temperature higher than air cooling. Or at least not for the reasons you mention.
Yes, watercoolers remove heat faster and more efficiently, but they still remove the same amount of heat energy. If your PC pulls the same amount of power from the wall, the same amount is going to end up as heat in the room, regardless of cooling solution. It is just going to do this by removing the same amount of heat, but at a lower equilibrium temperature (or Delta T).
That being said, better cooling usually means more overclocking, either by TurboBoost/GPU-Boost technology, or manually by the user, which increases power draw, and of course the heat being produced in the room.
Dude when you remove heat faster and more efficiently that means you're removing more heat. I think you've confused the heat produced with the amount of heat transfer and the temperature, and while heat produced will be the same for both systems in the same conditions, not all of that heat will translate into higher room temperatures since it's still somewhat isolated within the case. The more efficiently your remove the heat and dispel it outside the case the easier it's able to increase the ambient temps.
Sir Gildo So basically the equilibrium temperature of the room will be a bit higher with watercooling since heat enters the room from the case faster, but passively leaves the room just as slowly
", not all of that heat will translate into higher room temperatures since it's still somewhat isolated within the case."
You're an idiot and you're the one who's confused. Shut up.
You have two possible situations: Either the pc is heating the room, or the pc is heating itself. If the pc is at a stable temperature, no matter what it is, and it's pulling 500 W from the wall, those exact same 500 W will end up as heat in your room. Period. You can't avoid it. It's called thermodynamics.
The only way a water cooling solution can heat a room more than air is if it allows the cpu to generate more heat via overclocking and/or reduced throttling. Though it's possible in a room with poor insulation for there to be a steeper radiant as you move closer to the cpu, the actual total heat energy put into the room from the system remains constant to a specific power draw. I suppose you could argue the tiny amount of heat generated by losses in your cooling itself may contribute more than a lower power using system of cooling would through it's losses.
Indeed, christopher90, if a PC takes x watts to run, it will put almost all of that power into heat regardless of the method that heat is removed from the PC. I'm frankly a little surprised that Jay said that watercooling might heat the room faster, since he has had all other cooling-related arguments correct. We could nitpick and say that the pump in watercooling adds an extra tiny amount of heat (in the order of a couple of watts), but that is about a few percent of the total amount in idle and much less in load conditions.
Love the scene change for every myth. Youre produktion quality has skyrocketed in the new studio.
"Having water cooled systems is probably going to heat your room up even more"... WRONG
Jay, Jay, Jay. Man, so, so, so close.
What you're referring to (loosely) is the law of physics regarding the conservation of energy. That people still don't do their own research is why you're addressing this issue but basically, there is no additional energy being pushed into the rest of the room for it to be 'warmer'. It may very well exchange heat to the rest of the room faster, but the room will still retain/dissipate heat at the same overall rate. If your room evens out at 30C for an air cooled PC system in it, it'll do the same with a water cooled PC system in it, it'll just get to that temperature faster; much like, in fact, one of your other myths regarding loop temperature equilibrium.
So good work Jay, but your specific use of words here may very well have implied something other than what you intended. Please keep up the excellent content, and please consider the idea of designing/modding your own case for something truly unique.
Water cooling keeps the cpu cooler overall, so that extra heat goes to the room instead of the cpu, thus making the room warmer
@@tandlose That comment's over 3 years old, wtf are you even doing replying to it? Anyway, unless the room you're referring to is some sort of hermetically sealed box, any energy flowing into the room will escape out somewhere, so no, the room won't get 'warmer' to any greater degree between air and water cooling. It'll heat up *faster*, sure, but it'll level off at the same peak temperature when equilibrium is reached between and .
@@daiphelion7002 lmao 3 years ago
Thank you for the humor. Please bring back "Transitions", they were a fun touch. - Thank you for explaining water cooling to your fans and viewers. - Keep up the good work. - BTW, this is a water cooling video, however, my Noctua NH-U14S is a great CPU air cooler.
For some reason I do not understand there are people who still believe that Mac Pro is better than a regular computer in the same price range in video editing. I would love to see you make a comparison between the two
Because they are stupid. I saw a $1100 mac a month ago with a 1.9ghz duel core CPU, integrated graphics. 4gb of ram, a 1tb HDD, and a 1080p screen. Just because it's apple doesn't make the shitty specs better than a pc with a 7700l/8700/ryzen 7 and a 1070. Editing computers still use the GPU and CPU so having good ones is not exclusive to gaming. At best a mac has a 4gb AMD card at the price range where it should have a 1080 TI or 2 of them. Apple fanboys seem to think the apple logo magically transforms their low spec hardware into some ultimate editing computer.
forza1sra they do have an amazingly efficient OS i still heavily prefer Windows but just saying
They used to have hardware in the cpu that helped with this sort of thing. however with the switch to x66/amd64 intell chips they lost that advantage. They can get some gains with a tight optimization of software to known hardware configurations that is harder to do with a more varied ecosystem like you see in the rest of the pc space. However there is only so much you can get that way. So for an identical spec system they might be a few percentage points faster, but if the same money can buy you a significantly faster hardware spec, then the value quickly shifts away from them.
@@drieshuybens4594 so u would pay 1100 for just an OS. Thats like buying a ferrari for the looks but the engine they gave u is for a camry.
Hey Jay, just wanted to point out that the reason why fluid, particularly water, gets charged is because water molecules are polar and they can dissolve ions. De-ionized water does not possess any net charge (although it is still polar). As very small metal particles fall off from either the cpu or gpu block, they get dissolved into water and thus water gains charge capacity. But ions dissolving doesn't happen with any fluid, just those that are polar. In non-polar fluids, such as hydrocarbon-based fluids, metallic particles aren't dissolved. If the fluid is left to a standstill, these particles will eventually aggregate at the bottom of whatever is holding the fluid.
"Boils Down" you really don't want your loop to boil.
Digging the filming method with this one
*Spots 4K camera on the desk*
*Sets quality to 4K*
Oh, cool! :-)
Just a bit of complementing with some other details on the topic:
- There seems to always have been some discussion regarding tubing diameter and how that affects temps.
And the thing is as long as you're using good pumps (laing or go home) the difference in temps is pretty much negligible so this choice should be dictated by aesthetics and ease of access instead (also thinner tubing is easier to route).
- With regards on VGAs in series vs parallel.
If you have your VGAs on their own loop, pick what you think looks best. If you have them on the same loop as your CPU, you can use that to "shift" cooling capacity between your CPU and and your VGAs in a way.
Tl;DR: If you're more worried about your CPU, run your VGAs in parallel, if your GPUs are the ones you want to focus on, run them in series.
This is a rule of thumb and it has to do with the fact that water flow through the different components does affect a bit your temps. If you run your GPUs in series, your whole loop has a bit more resistance, but if you run them in parallel, the GPUs divide the flow flow between them. So flow for the rest of the system is a bit better but each GPU gets a less flow. So their blocks are performing a bit worse, but your CPU gets better flow instead.
Ultimately this is very much into nitpicky territory so to speak and the differences shouldn't really be too pronounced, unless you're relying on a weak pump.
- Bonus: pumps in parallel vs series
Pumps in parallel offer no benefit. Don' do it. Pumps in series are useful. You get better flow and redundancy. A stalled pump (true for DDCs and D5s) won't really hinder you much either
Are two 240mm rads nore efficient than one 480?
*more
Very interesting question though.
I would say yes.
2 240mm rads should have more copper than 1 480mm rad, think about the ends of a rad where the fittings are. So you should have more surface area to transfer heat (Very small difference though).
But there is more things to consider, the placement of the rads will make a difference. So if you can mount 2 240mm so that they only get ambient air (Probably easier than mounting a 480mm as an intake, never seen a case that supports a 480mm intake rad) then they will out perform the 480mm.
Would be an interesting to see a video on it though.
Depends of fins per inch and surface area
7:00 Loop order does affect temperatures. By a couple degrees C at most, typically, so one could say it doesn't matter. But it does make a difference. Any block coming right after a radiator is going to have cooler water flowing through it, which means heat transfer into the water will be faster, due to the increased delta T.
The one component that most stands to benefit from that is the CPU, due to the high thermal density of the interface. So putting the CPU block after all radiators might net you a couple degrees off top temperatures. GPU's are much easier to cool due to the much larger die area and much lower clock speed. And with all radiators together, the water entering them will be the hottest it is in the loop, making heat transfer out of the radiators faster.
Water temperatures never reach equilibrium across the loop. There are always gradients, where cooler water flows out of radiators, and warmer water flows out of blocks.
Do a video using mineral oil instead of water in a loop.
Mineral Oil would be too much dense in just a regular water loop. The viscosity would make the oil move too slowly in the tiny grill of the Copper plate that is in contact on the CPU
I have been using H2o and Propolene Glycol for years with zero growth (1 year between flush) goop or corrosion, I'm now trying H2o, Glycerin and silver dilution in my hard line loop. So far so good, keeps the pump quiet and seems to cool fine, Glycerin has a higher heat transfer rate/capacity than Glycol and is still antimicrobial.
I think its time you used a thermal camera and benchmarked different loop orders
What about just looking at the system temps lol
@@jeremymcadam7400 I like the thermal camera idea better. I think it would be really neat to actually see the heat travel through the loop and get dissipated until equalization
@@SirChaosS yeah fair enough, that would either prove or disprove the equalization theory
You should be able to go lower than room temerature if you spray water on the heatsink.
In theory the air passing by should evaporate the water quickly, sucking energy out of the system, that might be a fun one to test.
But most importantly:
MILK IS NOT A GOOD COOLING LIQUID
80% correct. Not bad, considering the general lack of understanding.
Loop order matters, and your instruments aren't calibrated well enough to detect it. This can be mathematically proven, but I'll go simple to explain it.
Heat transfer through conduction is based upon the difference in temperatures between the two surfaces. The higher the difference, the faster energy can be transferred. This is why at equilibrium the input energy from the components equals the output energy to the surroundings. This means that the more energy dumped into the system at any one point, the less can be dumped in at the next point. Thus, for the most energy transfer (and the lowest equilibrium temperature) you want to organize components such that they output more thermal energy as they get farther from the reservoir out.
Why do most people think this is not the case? Put simply, your instruments are incapable of detecting it. The heat capacity of water is 4.184 Joules per gram per degree Celsius. This means that if your system has a total of one liter of water in it, or 1000 grams, it would take 4184 Joules of output energy to raise the temperature one degree Celsius.
Reviewing basic heat transfer. equilibrium would then be based upon temperature difference at the block and at the radiator. whenever energy transfer from block = energy transfer at radiator, dictated by the temperature of the transfer fluid, then you reach a steady state.
Now, what is a Joule, and how does it relate to watts expressed in a TDP. 1 Joule = 1 Watt/second. So, a system with a TDP of 500 watts is the same as 500 Joules, which means a system where the TDP is 500 watts would only change the temperature by about 0.125 degrees. Putting on our scientist hats for a minute, when you measure temperature in degrees Celsius how many times have you seen a gauge capable of 0.0625 degree precision? That's right, almost never.
If your system only outputs 300 watts (a decent GPU and consumer grade CPU) your temperature delta is nearly insignificant.
So, you missed the math. The math, and very precise sensors, could show the difference. Additionally, a different fluid (mineral oil has a capacity of about 1.67 J/g/C) would show more dramatic results. Water is functionally magic when it comes to fluids, which is why people don't quite understand what they are seeing. Please stop the disinformation. If you want to debunk this, the correct response is "Given available and affordable sensors, you won't be able to detect any appreciable difference in performance directly related to loop organization." There's a huge gap between there is none, and I cannot detect one.
Thought experiment. You have a fire at one block, a small resistor at the next block, and a radiator. By organizing them such that the resistor is first, then the radiator is last, you can transfer energy from the resistor out of the system with a sufficiently large radiator. If the fire is first, then the resistor will actually absorb thermal energy because the delta in temperatures will always favor heat transfer from the loop fluid to the resistor.
Parallel vs serious loop? 6:56 :)
Peter Kostovčík serial
Cereal
Thanks Jay great video, i was interested in the bit about the length of time it takes for non conductive liquids to become conductive, it would be cool to see a follow up about that as well. Thanks again keep it up.
can I use water from cars cooling system( radiator )
Yeah, Luke from LTT done this in Scrapyard wars (season 2).
I think the only reason people don't use it would be because of the colour(s) of it. It should be a superior alternative for long term stability as it is designed to last for years in a system do extreme heat cycling with several different metals in the system.
If you dig deep enough there's at least one manufacturer that makes a product that does not use water as a mix, so no evaporation, no boiling etc. Evanscoolant.com
esmailjam You can, but there's no benefit compared to regular distilled water with corrosion and growth protection. Cars are supposed to have their coolant flushed every two years, and the same is recommended for pc cooling loops. The only case in which using automotive coolant makes any sense would be if you were running your pc cooling loop in an environment where the ambient temp is below freezing.
I believe it comes down what antifreeze it might have in it since it can eat the orings or something along them lines.
You can, but glycol actually has less heat capacity than water, meaning it would be functionally equivalent to using a smaller fluid capacity (which, as Jay said, may not matter). More importantly, the glycol will completely eat some kinds of tubing so unless you built it yourself and know every plastic type in the system, I would not recommend putting glycol in your cooling loop.
I would be very interested in seeing numbers of how many radiators are needed for different setups if you want sub 1000rpm fans and working towards silent systems.
Public service announcement please stop saying first
Bad no
Streets Speed 484 first
Tucker Bergantino it's not hard, try starting a discussion instead about the video or just something else
PENIS
Streets Speed 484 FIRST
2:55 Completely wrong. Assuming that the CPU and other blocked components put out the same heat regardless of cooling method, that heat is going into the environment regardless. The room will heat up by the exact same amount over the course of the day.
That assumption isn't safe, however. Transistor switching is sensitive to temperature (electrical resistance goes up with temperature). At a given clock speed, a processor at a higher temperature will consume more power than a processor at a lower temperature. Cooling the CPU and GPU more effectively will reduce their power consumption for a given clock speed. That means less waste heat, which means a cooler room. How much cooler depends on a number of factors, and the contribution of the computer is likely to be swamped by other factors, but it might just be measurable under the right circumstances.
"(electrical resistance goes up with temperature)" -- That's true of conductors (i.e. metal) but not true of semiconductors. Semiconductor resistance decreases with higher temperatures. That's why semiconductors without thermal protection are susceptible to thermal runaway. Higher temps->more current flow->more heat generated->repeat until meltdown.
Jay cmon dude.... :-/
1) Wrong, a larger reservoir gives more thermal mass in the system. If the radiator can support the cooling function, it will give you "lower temps" because you aren't heating the total mass the same amount. Q=MC(T1-T2); MASS is a function of thermal power. Your transient response time will also be slower proportionately to the amount of mass gained.
2) At the same fan speed, larger radiators will have a faster thermal transient response; albeit not below ambient, and point of diminishing returns applies
2b) wrong, heat in, heat out. If youre cpu is putting out 150w, its 150w. doesn't matter if youre cooling it with air or water; youre still putting 150w of heat in your room.
3) Certain chemicals don't strip ions; some of the nano based fluids don't strip ions because they dont have water in it. Think of mineral oil cooled pc's. The only ionization that will occur will be from friction stripping.
4) Dye's are so all over the place; youre right about this one
5) Right on.
Yea, I can put a cup of water on the stove, or a gallon of water on the stove. Sure, eventually they'll both reach 100C but it'll take so much longer for the gallon to reach 100C that in practical use like gaming it's literally going to keep your temps lower.
he's actually correct, regardless of the amount of fluid being used, eventually, they would both equalize at the same themp, sure the larger amount would take longer, but it would still reach the same temp. Ran this test in my thermodynamics class years ago.
They are arguing that if they have a larger res it takes longer for the system to reach equilibrium. Which, they're right, it will take maybe a second longer.
Thermal transfer is faster across larger gradients. A larger reservoir at most creates a small delay in reaching equilibrium.
@Jg. You're wrong. He's right on #1. #2 he's right again. Water cool is going to be better at moving heat from the component to the ambient air. Why would people water cool if it didn't lower the temp of their components. If you air cool to 90C but water cool to 85C you have 5 degrees of temperature moved from your cousin to your room... Hence warming the room more that air cooling
actually what jay says at 2:54 is a myth: The amount your room heats up depends only on the total power consumption of your system (law of conservation of energy). The only thing that makes your room hotter when watercooling is the additional power your pump (and maybe additional fans) need, which most likely is negligible. In other words: a 1000 watt drawing pc will heat your room exactly as much as a 1000 watt heater, no matter how you cool them.
Loving this content, one I've heard is that having your pump run faster will give you lower tempretures
what would be really good for debunking the loop order myth would be to have some thermocouples installed at various points around the loop. If what Jay is saying is true then each thermocouple should read nearly the same temperature on either side of each waterblock.
You could also do some quick math using flow rate(assuming you could determine that), specific heat and the wattage of each component under a waterblock to find the theoretical increase in temperature over a waterblock.
I just calculated the temperature change over a 500 watt component would be about 3 degrees celsius.
That was assuming a flow rate of 0.05 L/s and a specific heat of 4180 J/kg*K.
So yea, Jay is right. Not a big change in temperature.
Can't wait for the loop order video. Have some friends who still don't believe it doesn't matter. With that video I can finally shut them up... Great work Jay.
Myth one: watercooling has reasonable prices.
Really useful info. I've never built a liquid-cooled system, but I want to do this for my next build!
Love the different filming spots for each myth. Was almost hoping the last one would be from the toilet or something.
I am loving the videos to remind us of the basics and now even more so since I will be doing a new build and want to watercolling because the system will have top of the line components and I want the most i can get out of them.
This video is a prime example of why physics should be a mandatory class in high school.
For those curious, you can calculate how much heat your radiator has to dispell using the following equation:
Q = (k * A *🔺️T * t) / L
Where "Q" is the heat transferred per second, "k" is the thermal conductivity of the material (which is 205 W / m * K), "A" is the surface are of the material, "delta T" is the temperature gradient (most commonly 50 °C compared to 22.2 °C), "t" is the time (1 second in most all cases), and "L" is the thickness of the material (generally 120mm or 0.120 m).
Obligatory "Do not attempt" Warning: Do not attempt any modifications to your computer unless you're a trained professional or have extensive knowledge in thermodynamics/electrical engineering. Seriously, you can and will damage your parts.
I wonder what the lens he's got on that rig on the left at the end of the vid. Honestly a filming kit video would be super cool
Hey Jay. I love your new studio and all. But you should do videos at the house more. I miss seeing skunkworks in the backround and the more down to earth vibe your older videos had. Don't get me wrong I love the step up you've taken with your channel and the content has been getting better and better but every once in a while it would be really cool to see a video at home for old times sake😉
I love the scene changes. Jayz getting serious about this.
Can't wait for the myth buster episode!
You should do a video about 5 myths regarding Tech UA-camrs. I am sure there are a ton of them out there and it would be cool to hear you address some of them.
Jay loving the background scene changing in the video
Please test static pressure fans on case, are they better or worse than airflow fans? thanks in advance. I love your unique and useful contents! :)
Have you ever seen any reviewer do something like investigate transient responses and used them to estimate the cooling power of radiator/fan combinations? Like, let the system heat up to a steady state then cut the CPU load, and then for a range of radiators and fan speeds, from how quickly it cools and knowledge of the mass of fluid in the system and the specific heat capacity of the fluid you'd know how much heat they could dissipate and actually be able to compare apples to apples between brands.
I love the scenery changes! Makes it way more enjoyable!
A few weeks ago I was browsing through Facebook ads, and a tower with a custom water loop popped up for a reasonable-ish price, so I looked at the pictures to get a better look at what condition it would be in. I tried to follow the loop through by zooming in and I figured out that it was in the opposite way. More specifically, the pump was pushing water to the radiator, which then sent it to the components, which ended up in the water tank. Basically the water was hitting the radiator BEFORE the components it had to cool down. I really wonder if the guy even managed to sell this thing.
3:00 "A watercooled system will heat up your room even more" This is incorrect apart from the extra energy consumption from the pump. If the room is a closed room the amount of energy consumed by electricity dictates how much the room heats up in total. Even if the heat transfer from the case to the room is lower with regular fans, it will still over time affect the temperature of the room. One could maybe say it will heat the room up faster while the PC is running, but not "more"
Well the pc would heat the closed room up to what ever temp the air is coming out of the case or rad (normally around 40-60c depending on current use). So in a way it would heat the room up more then you prefer it to be.
@@pheelix- That is not entirely correct. eventually there will be heatsoak and the room temp=rad temp. This will prevent the CPU from getting cooled properly over time. Thusly increasing the temp of the CPU, then the temp of the water which will raise the temp of the rad which will raise the temp of the room. Whent the CPU hits T-Max it will shut the system down and the room will cool :p
@@savagedk True, but would the system thermal throttle itself to the point of equalization before reaching T-Max and shut down?
@@pheelix- Good question. Maybe Jayz will test that some day just for fun :D
If T-max is higher than 100C then I take it the cooler would explode due to pressure at some point lol.
@@savagedk I doubt he would test it. Probably to busy playing with all his free stuff.
I am so happy for you that you hit 1 million subscribers! I have been rooting for you to get to 1 million! Congrats!!!
Propylene glycol I feel might be the magic cooling fluid that everyone is searching for, it provides corrosive protection, its molecular which means it should be non-conductive and should have good thermal conductivity. Would love to see a test bench please. Please try a leak simulation by spraying some on to a old system just to see what happens, try it straight then 60/40 with distilled water. Can use food grade PG which is used in E-cig fluids or Amsoil makes a nontoxic antifreeze which is industrial grade PG. Thank you.
It's been a long time since I studied any of this in school, but from what I can remember...
You're not heating up the room any more or less. Heat is energy and it never turns into something. Your components produce a certain amount of heat and that is the amount of heat your computer produces no matter what. The only factor there is how well your case moves it out of the case via the fans, but having too little airflow will heat up the inside of the case as you don't move it out of there efficiently enough, but since it can't turn into nothing the heat will always seep out into the room - it's a matter of how fast/efficiently. Moving the heat/energy into the room will then enable moving it elsewhere. Energy is eternal.
I have a 1080ti and a 5960x both cooled on water with 2 Bitspower 480 x 30mm radiators and my temperature difference on AIDA64 under load is 3 degrees whether my fans are at 800rpm or 1500rpm. So the extra radiators help a lot with allowing the fans to run quietly and efficiently. My single 240mm AIO on the same setup would run about 15 degrees warmer at idle and would have about a 10 degree delta between those same fan speeds (same exact fans as well Corsair ML120 Pro.)
When you will make the video concerning the loop order, can you please add some information on how do bend and use hard tubes for beginners? That would surely be very useful to see some of the bending tools in action.
Really good video in comp to the common mistakes as changing the environment for each fact is just gorgeous!
Hi Jay, please include pump/rez into rad then into blocks, i was under impression, the pressure hitting the water blocks generate turbulence for better heat dissipation