Couple points! Thank you Nikolay Pelov and linuxgeex At this low a voltage, the motherboard should never arc. Also, the true thermal barrier isn't even the increased resistance. It's the reduced impedance due to increased quantum tunnelling as temperature goes up, which increases leakage current, which is the majority source of the increased heat once you get into thermal runaway.
It’s funny, at the beginning he reminds me of an employee I’ve had for 20 years- the guy is nearly 50 and still a virgin… Once Linus was gone, though, he showed a little more personality.
One of my favourite memories from my undergraduate is completely forgetting the project we were supposed to be doing and dunking everything we could find into liquid nitrogen, including a USB stick which we then transported via tongs to the computers and plugged in 😅
@@yaago - yup! i'm no content creator, but it been dawning on me just how much time other people sink into making videos. it definitely aint trivial lmao
Moving is very stressful. Moving into a mansion with a bunch of employees to help you and a bunch of cool life-goal tech stuff to setup is significantly less stressful. Might even be therapeutic.
Contrary to popular belief, big dudes are actually pretty chill. You go to any gym in the world, approach the biggest guy in there and talk with him and he'll 90% be the chillest dude there. Strength makes you kind
@@zhongxina9420 a buddy of mine is 7'2". Dude's nickname in high school, from football, was "The Fridge". He's the most caring and chill dude I know. So in my experience, your comment is the truth.
I have to say this, I am really impressed by your water bottles and how well they are isolated. I never thought they can hold something like liquid nitrogen for such a long time and that you can hold it with your bare hands.
@@YannickBo For some extra convincing, if I fill my water bottle with ice and ice cold water and leave it shut, it won't finish melting for around 20 hours.
Oh men, When Luinus comes back and is amazed 13:29 LOL, I know that he is not faking it. I honestly feel the same way just by looking at this video, soo well produced with a touch of love, personality, and a real passion for all of you guys' disciplines. Thanks for soo much, LTT Team.
Seeing a guy that looks more like a MMA fighter than a computer geek doing some serious overclocking to be considered one of the best overclockers over there is just refreshing to watch.
im working with cryogenics for work and the samples we are working with need to be handeled in LN2 at all times so seeing this i got some ideas. If you put the motherboard in a big container, the gas nitrogen (from the boiling) would fill it because it is heavier than air and you would not have air (which has all the moisture) around it. If you have a university or research institute around, you could think about contacting them. Cryogenic work is quite common nowadays and there is probably some experts around that can help with this
These guys handle some ridiculous materials. They do some ridiculous electrical mods. They engineer all sorts of ridiculous hacks and workarounds. But they insist on using "common" consumer parts in "common" consumer conditions. That means a computer on a table in normal air, no complete immersion, no vacuum, no exotic environments. Dunno why, exactly. The hobby is cutting-edge math, science, and tech on one hand. It's unscientific bumbling on the other. It's ruthlessly competitive on one hand. It's team collaboration on the other. It's methodical experimentation and calculation on one hand, it's reckless gamble on the other.
To add to Groentje92 comment: Make the container out of 30mm plexiglass then enclose the top. whole thing rubber sealed. Add a flask with a tap to allow the topping up of liquid nitrogen. Add a vacuum pump. this will allow the chamber to be pulled down to a lower pressure, lowing the boiling temperature of liquid nirogen, allowing even colder temps. Also might be worth adding a 45psi burst disk,.. just in case of over pressure even if vaccum pump fails at some point.
A video like this has been a long time coming. As a fellow LN2 overclocker, I’m happy to see this. While there are a lot of caveats in some stuff you said, you did an excellent job in packing it all in to a short time-frame. Kudos.
Speed of light is also a big limitation, 6.9 ghz means light can only travel 43mm (1.7 inches) per cycle, which is just insane. The size of the die is 20.5 mm × 10.5 mm, now that clock cycle is only within one core inside that CPU die, but still insane, you'll start to see why there isn't a much of room left.
Also the speed of electricity is slower than the speed of light, which could very well get it to within the range of the die if you tried to get much higher.
Gonna be fun next gen with rumours Zen 4 is hitting 5.8Ghz in the lab in early silicon samples, we're gonna have many more attempts at very high clock speeds coming up.
@@batt3ryac1d All CPUs can be both inefficient and efficient, it only depends on what power they run them at from the factory. If you're worried about efficiency just undervolt it like I do.
I was having a heart attack throughout the video, Linus has a reputation for dropping/knocking things over, imaging that happened when he's pouring that liquid nitrogen.
@@bobuccman1424 Nope. I actually was burned by LN2 before, and while the Leidenfrost effect works it doesn't actually protect you from nasty burns :/ Especially with a whole thermos full of it😅
watching Jay do liquid nitrogen overclocking was really fun so it's good to see you do a video like this! i wouldn't ever try this as i just have no desire to but it is enjoyable to watch
Just an FYI (months later) you'll probably encounter the leidenfrost effect with liquid nitrogen. Effectively the liquid boils away and shields the core from cooling; this is why you can dip your hand in liquid nitrogen. If you convert to a pumped/flow setup you will avoid this. It would also allow you to use other cases like helium with little change. Source: in grad school we worked with cryogens and temperature controllers for experiments at 4K or below
@Linus the true thermal barrier isn't even the increased resistance. It's the reduced impedance due to increased quantum tunnelling as temperature goes up, which increases leakage current, which is the majority source of the increased heat once you get into thermal runaway. That is what will damage the CPU, and that's the reason for throttling. Well... unless you're throttling before you melt the board's power stages or power delivery traces. We're at that point now where some mobos will fail before the CPU!
yeah it sounded wrong to me since semiconductor should have lower resistance as temperature goes up unlike metal. Thanks for clarifying the whole thing.
Ah, this is the post he referenced at the top! Thank you. Now I know that silicon has a negative coefficient of resistivity, which I think explains why he said that the board doesn't like to power on when it's cold- The impedance of the silicon is too high when it's cold & doesn't allow enough current flow.
@@nickpayne1182 At the frequencies they're pushing, 99% of the chip is a series of connected RLC tank resonators being coerced into operating literally in harmony. For best efficiency those little tanks are designed to operate happily in a given frequency range with a given amount of power based on capacitance, inductance, the driving voltage, the driving frequency. The rate of change of voltage is based on the resistance (metal and semiconductor) and impedance (metal and semiconductor.) The resistance of both goes up quite linearly with temperature. Impedance on the other hand is quirky, lol. As the capacitance and inductance deplete they induce current flow which increases impedance to the depletion of the inductance. Higher temps decrease that impedance but prolong it. Net effect is that the transistor takes a little longer to switch, and so requires a higher voltage to switch fast enough to keep up with the clock rate, hence why cooling and under-volting allow you run higher clocks at lower power. Quantum tunnelling does the opposite. It's behaves similarly to how an avalanche starts thanks to vibration. The electrons want to get from A to B. Semiconductors can be biased to operate as either conductors or insulators. Insulation works to prevent electron movement by presenting a uniform wall of electron shells perpendicular to the desired path of the electron(s). As the insulator's molecules heat up and shake larger gaps intermittently appear in that wall. When the gaps get large enough, the odds of electrons jumping through them increases (quantum is all about odds lol). That shakes the molecules even more, and soon you have an avalanche of electrons migrating where they shouldn't, plus more heat than was designed for. If the heat is sufficient it can break molecular bonds and the electrons will start pushing atoms around, aka "electromigration". CPUs have heat sensors to make themselves throttle before that happens, but those sensors can fail and allow a CPU to self-immolate, sadly, and that's part of why TDP exists - the CPU will also throttle above a certain power usage window as a backup failsafe anti-suicide measure. Intel K series processors have that safety feature turned off... you pay for less safety lol.
@@linuxgeex Holy cow, thank you for the explanation & giving me a good starting point to do more research. I learned the basics of electrical theory in my apprenticeship, so luckily this isn't all Greek to me. Your explanation sounds very similar to what I was taught about reactance & hysteresis in AC circuits. Thank you again for the explanation.
Damn, Splave ole buddy, recognized you right off - looking good Bro! Teammates years ago on the "No Borders Extreme Overclocking Team". He was the best way back then and just kept getting better!
a cool extention of something like this could be to put the whole system in a pressurized nitrogen atmosphere so moisture isn't an issue then use a cryogenic heat pump similar to the ones used to make LN2 as a cpu cooler by maybe using LN2 or helium as the cooling liquid. have a raspberry pi or the like managing flow rates and temps and it could make one hell of a server (wouldn't want a set up that big on a desk so in a rack it must go)
The problem is, the action of the LN2 boiling rapidly is actually what sucks out so much of the heat. It pulls a lot of thermal energy out from the pot to go from the liquid to the gas state. This is actually very much how freon(r134) in our AC systems works, the transition from liquid to gas, and back, makes it able to absorb and release that heat so efficiently That's a really neat idea tho, I wonder if there is a way to implement it in a moisture free pressure vessel that still allows the energy transfer as efficiently(maybe more?)
You can call it a good day when Linus doesn't lose a hand while messing around with liquid nitrogen. I'm pretty sure Yvonne only found out about this AFTER it was already filmed.
@@clavius5734 Gloves are actually bad when handling LN2. Your hands are hot enough when it hits them the LN2 boils and you're protected by the vapour creating an insulation layer (Leidenfrost effect). Put on gloves and the LN2 soaks in leading to v frosty hands. -AC
As was said, you REALLY don't want gloves or even socks when using LN2 for OCing. LN2 isn't really THAT dangerous assuming you know what it does. Movies make people think it'll freeze you instantly and that's just false.
To give a little more context as to the dangers of LN2, it has 2 primarily. First is its displacement of oxygen causing asphyxiation (don't breath the gas). Second is its cold, but when touching your skin it due to the temperature gradient causes the LN2 to flash directly to a gas with a sort of barrier forming between your skin and the LN2. If you put on gloves this doesn't happen, it soaks into the gloves and super cools the gloves leading to burns from it. This will happen with socks and other fabrics too.
This cool guy is like when you go see a car mechanic, big and quiet, but knows their sh*t inside and out, fantastic vid, one of best you have done lately :D
Another possibility for potentially avoiding condensation altogether would be to do what you're doing here with the extreme overclocking inside of a walk-in freezer. The cold air tends to be dry and has the bonus of keeping everything cool, but not ridiculously cold. The LN2 would still boil in a walk-in freezer. You just may need to make use of a heat gun or that torch a little more. If you did this inside of a sub-zero environment, you wouldn't need to go to the trouble of building a vacuum enclosure like the one I described in my previous comment. I'm sure you could find someone who would let you borrow their walk-in freezer for a few hours. ...Or you could possibly even buy your own if you wanted. You could maybe use it to continue working towards breaking and even setting your own overclocking records. 😁
You would need a compressor and a condenser to get the vapour back into a liquid. That might require a stage or too extra (like an LN2-to-ammonia heat exchange and a second compressor loop).
@@xuser48 Apparently, liquid nitrogen is considered impractical as a looped coolant, but CO2 is actually used in commercial refrigeration. Apparently a bit expensive, but I read that it was being marketed as an environmentally friendly alternative to the currently popular refrigerants. It was interesting to see it suggest big systems might leak hundreds of pounds in a year.
You guys should try a raised reservoir feeding into the pot governed by a valve that you can modulate the drip rate to keep the rate of cooling/coolant use more accurate.
Only time I've messed with liquid nitrogen was with overclocking a old socket 754 sempron i think 2800+ which was stock like 1.6 ghz pushing it too 4 or 4.1 ghz On a poor matx gigabyte board with an nforce chipset. At university ca 2013 or 2014 with a ocz 650w powersupply hooked up too it i think. Agp slot on the motherboard & integrated graphics was first used but that stopped showing a display after a bit (go figure lol.) So a random agp card was used instead. Ram was like one or two sticks of 400 mhz cl2 ram at 512 megs or 1 gb in total. Pointed a fan at them & fan at the vrm which was completely bare btw, no heatsinks at all on either of them. Classmate got too work building probably one of the most ghetto liquid nitrogen pots ever... Step 1: Regular size pringles can, don't remember if it was either sourcream or a regular one. Step 2: Gobs of solder on the bottom of it too raise it up around the motherboard a bit & get a sort of cold surface plate, think it was leaded solder, might have been unleaded or a mix of both. Ofc making sure not too melt a hole through the pringles can bottom which is kind of aluminium foil ish so close enough right ? Which also didn't crack from the cold & worked surprisingly well after some sanding too smooth it out flat which took him probably like 35 minutes or so lapping it fairly smooth too the touch & flat. + ofc thermalpaste later used for better contact. Think it was MX-2 or MX-4 & caked the cpu in it. Step 3: For mounting the pot somewhat more rigid then gravity, 4x holes where i think wooden sticks or pegs threaded through, zipties through the backside of several mounting spots on the poor motherboard too tighten it down doing it gently a bit at a time too keep it sticking down somewhat squarely on the cpu while making the mounting tight against the cpu. Ghetto as all hell but worked surprisingly well enough too mess with & get it too boot into the bios at 4 ghz after all, probably not an overclocking record by any means but fun too do anyway. Also filled the pot with bottlecaps & other random metal garbage from the workshop & a box of random pc screws too absorb & keep the nitrogen from dissipating too quickly, we surprisingly due too the Jank mounting we did had nearly 0 issues with issues booting up having readings at like -102c or so, but might have been the sensor just completely freaking out as well. Ofc we used paper towels & a bunch of other stuff too insulate the board from any moisture the best we could together with the fans. If we had a pci ide interface card we probably could have managed too make the poor thing boot into windows but the onboard controller & probably most of the chipset was noping out after like 2.6 ghz already & we kept upping the fsb as long as it booted into the bios after maxing out voltages on chipset, ram & cpu lol.
Ive reached 4.1Ghz or 4.15Ghz too on a sempron; using a phasechange keeping the CPU at -45 on avg. Was a fun day. Installed the phasechange later ontop Ati 2600 pro which had stock clocks of 600Mhz, soldered all caps with larger or better ones, installed a external PSU instead of being fed through the PCI-E Bus, and managed over 1260Mhz core clock lol. Once i turned the voltage way beyond 2.2V the sparks just flew out of the back of the card and the party was over.
And also (and this is maybe the 5th time I write this) you CAN (and should, it's not difficult) put your motheboard in a controlled atmosphere (i. e. a box with a Argon or Nitrogen gas flow) so there is NO MOISTURE in the air. In the research world this is done for example to protect samples exposed to radiation from oxidation.
i love how splave's face glows red like he's a little kid going over to his new neighbor's house and seeing that his neighbor also also likes playing with g.i. joe action figures. its so innocent and wholesome 😊🤗
If you remove the air and turn it into a vacuum like out of space, wouldn’t the metal components weld to each other and the liquid turn to jelly and evaporated?.
I would guess the latent heat energy absorbed when the LN2 boils away is important for keeping everything cold enough, and a vacuum makes the boiling point too low
You know what else helps? A well sealed room and a dehumidifier to pull more of the moisture out of the air. Won't get it all but if there are any holes in your moisture barrier it will reduce the risk of anything getting too wet
LN2 is pretty damn cool (pardon the pun) and it looks pretty fun to overclock a CPU with it. Liquid helium on the other hand, although it's even colder than LN2, does annoy me when people waste it on just overclocking a CPU... Helium is a rare resource and is something we're quickly running out of. The more important uses are for cooling magnets in high power particle accelerators and MRI machines - i.e. stuff that really matters.
:( but I really like watching balloons go weeee up up and away. Lol also makes u talk like your talking into a fan now thst is what I would call important
@@anthonycote2916 IIRC It is quite abundant on a scale beyond our planet, but being a extremely light gas that doesn't form heavier compounds like hydrogen does to form water it gets stripped away from the planets atmosphere from radiation. I only did a quick search so this may be wrong but helium and hydrogen have similar concentrations in our atmosphere, but since hydrogen gets trapped as water we have a much larger supply of it.
Also, I remember at school we had some scientists form a lab came in and show stuff with nitrogen. I remember they froze waffles and chocolate for a second and then gave us to try it.
Hmmm, how cool would a Novec submerged PC with a LN2 powered heat exchanger to pull heat out of the Novec be? I think Novec 7000 boils at 34° C but freezes as low as -122° C. Submerging the test bench in Novec would remove any condensation risk. It’d also likely stay colder way longer since Leiden frost wouldn’t be an issue, and a few liters of Novec at -100° C would be a massive thermal sink.
1:33 - we used that trick to quickly "sweep" the lab floor! If you pour nitrogen onto the floor, it propagates in a "wave", carrying dirt and litter until it hits the walls. Then you can just sweep it. Kinda like vacuuming, but much more fun.
Couple points! Thank you Nikolay Pelov and linuxgeex
At this low a voltage, the motherboard should never arc.
Also, the true thermal barrier isn't even the increased resistance. It's the reduced impedance due to increased quantum tunnelling as temperature goes up, which increases leakage current, which is the majority source of the increased heat once you get into thermal runaway.
Is this linus's personal youtube account?
Okay did not understand it at first
please make pc inside the vacuum chamber
ok
Nice It Linus Jobs!
Allen is an absolute UNIT
I guess he does only 2 things in life, lift and OC
He OC CPUs and muscles
"Look at those gains! +1237 points in R20 compared to the run before!"
Dude is like, three or four absolute units.
"Every time I bluescreen...I do ONE push-up."
He benches the 230L of LN2 cylinder while waiting for cinebench runs
rumours are is that Splave didn't use nitrogen to cool the cpu, he just told it what he wanted and the cpu was too scared not to do anything else 💪😝
that what gains are for 💪
Absolute unit of an overclocker
Loved this guy. He's got a gentle giant vibe to him
It’s funny, at the beginning he reminds me of an employee I’ve had for 20 years- the guy is nearly 50 and still a virgin… Once Linus was gone, though, he showed a little more personality.
loved?
The big kid from Camp Onnawanda
a giga chad at its finest
@@ncrawford1488 lol bro what - keep that to yourself next time. also I bet he still gets more play than you
I really like Splave. He's so chill and knowledgeable. Though, to be fair: most of Linus's crew has a unique and fun personality.
Hes not the crew
@@pmudriBaki?
Splave is a massive guy. If I was a CPU I would simply overclock as high as he wanted me to
Real
Sus
LMAO
#gay
@@0xBlez YEAH
Linus got jealous of Steve's giant well polished cylinder.
of the what
ayo
:SUSSY:
Lmao
:(
P-cores at 6.9GHz and E-cores at 4.2GHz. We're hitting all the nice numbers in this video.
ないす 남벌쓰
@@ADeeSHUPAwhy dont you like it
@@M4D0GG0 私は決してあなたを離れず決してあなたを失望させず決して逃げず決してあなたを見捨てません
@@randomgamingin144p did you just fucking rick roll me in Japanese?
@@M4D0GG0
ないす - Naisu - Nice
Splave is the sorta dude you'd want having your back in a rough alley and some sort of zen tournament of absolute controlled power
Do you want this guy in your back?
@@ramasso 🗿✋
Was “zen” a pun because amd? Im overthinking this..
A tournament of power
Alex and Splave just doing it perfectly, completely chill, no problems at all.
_[Enter Linus]_ Wiiieeeeeeeeeeeeee! Whooo-OH, BSOD (15:23 15:39)
I sincerely hope that if a CPU manufacturer ever releases a SKU that can hit 6.9ghz, they’ll call it the ‘Nice’ edition.
N1C3 model name
it better have a 4.2Ghz base clock
Intel 6969NICE
The Giggity Edition
Intel 6960 Nice Lake edition
One of my favourite memories from my undergraduate is completely forgetting the project we were supposed to be doing and dunking everything we could find into liquid nitrogen, including a USB stick which we then transported via tongs to the computers and plugged in 😅
Did it still work?
SPLAVE needs a TH-cam channel. Dudes vibe was amazing. Just a joy to watch
📩📩☝🏻 ..
..
if he makes a youtube channel he will loose the time editing videos uploading etc therefore his ability with cpus would decrease
@@yaago - yup! i'm no content creator, but it been dawning on me just how much time other people sink into making videos. it definitely aint trivial lmao
@@yaago he could always have somebody do the editing for him
@@gardensofthegods not everyone wants a social media presence. Especially a video channel.
I love how much fun Linus has been having lately.
@@samusaran6936 Why not?
@@samusaran6936cuz he’s doing what we all wish we could do, are you just jealous ?
Love your job and you will never work a day in your life.
Moving is very stressful. Moving into a mansion with a bunch of employees to help you and a bunch of cool life-goal tech stuff to setup is significantly less stressful. Might even be therapeutic.
I love how the professional overclocker is standing there watching Linus explain what he does lol. He looks like he wants to jump in and explain it.
Imagine being this guy, having to be there listening to Linus explain what you do while making wiii, puuuu, piiiii noises.
He looks so intimidating with his arms crossed and the serious expression
It's like watching a kindergartner explain to another kindergartner what they just learned.
more like a "volunteer" overclocker than a "professional"
Almost like Linus is the presenter or something
That was the most chill dude I've ever seen. So soft spoken. So calm.
It contrast well with Linus going "WEEeeEe" while pouring the liquid cold
Contrary to popular belief, big dudes are actually pretty chill. You go to any gym in the world, approach the biggest guy in there and talk with him and he'll 90% be the chillest dude there. Strength makes you kind
@@zhongxina9420 a buddy of mine is 7'2". Dude's nickname in high school, from football, was "The Fridge". He's the most caring and chill dude I know. So in my experience, your comment is the truth.
I have to say this, I am really impressed by your water bottles and how well they are isolated. I never thought they can hold something like liquid nitrogen for such a long time and that you can hold it with your bare hands.
This video actually might be the final push for me to get some
This video is way better marketing for the bottles than anything else they can do.
you need a vacuum insulated water bottle to prevent the ln2 from freezing your hand
Tbf, even normal vacuum bottles work just fine, as long as they are actually vacuum flasks
@@YannickBo For some extra convincing, if I fill my water bottle with ice and ice cold water and leave it shut, it won't finish melting for around 20 hours.
You really exceeded our expectations this time, by bringing Splave, a world famous overlooking expert to hit Golden Sample CPU's with LN2. Good Job!
69 likes
*overclocking
@@kr19569 Maybe he's also famous for overlooking things
@@kaldo_kaldo haha, right on!
Oh men, When Luinus comes back and is amazed 13:29 LOL, I know that he is not faking it.
I honestly feel the same way just by looking at this video, soo well produced with a touch of love, personality, and a real passion for all of you guys' disciplines.
Thanks for soo much, LTT Team.
Seeing a guy that looks more like a MMA fighter than a computer geek doing some serious overclocking to be considered one of the best overclockers over there is just refreshing to watch.
im working with cryogenics for work and the samples we are working with need to be handeled in LN2 at all times so seeing this i got some ideas. If you put the motherboard in a big container, the gas nitrogen (from the boiling) would fill it because it is heavier than air and you would not have air (which has all the moisture) around it. If you have a university or research institute around, you could think about contacting them. Cryogenic work is quite common nowadays and there is probably some experts around that can help with this
These guys handle some ridiculous materials. They do some ridiculous electrical mods. They engineer all sorts of ridiculous hacks and workarounds.
But they insist on using "common" consumer parts in "common" consumer conditions. That means a computer on a table in normal air, no complete immersion, no vacuum, no exotic environments.
Dunno why, exactly. The hobby is cutting-edge math, science, and tech on one hand. It's unscientific bumbling on the other. It's ruthlessly competitive on one hand. It's team collaboration on the other. It's methodical experimentation and calculation on one hand, it's reckless gamble on the other.
To add to Groentje92 comment:
Make the container out of 30mm plexiglass then enclose the top. whole thing rubber sealed. Add a flask with a tap to allow the topping up of liquid nitrogen.
Add a vacuum pump. this will allow the chamber to be pulled down to a lower pressure, lowing the boiling temperature of liquid nirogen, allowing even colder temps.
Also might be worth adding a 45psi burst disk,.. just in case of over pressure even if vaccum pump fails at some point.
@Danko Kapitan structural biology. The microscope we work with works in cryogenic condition so we have to prep our samples accordingly
@@shanewilson3653 I didn't want to go into vacuum as it is probably alot more difficult to pull of but you are completely right. Woud be even better
@@shanewilson3653 when tou close it off you cant heat the cpu up anymore to boot 🤔
A video like this has been a long time coming. As a fellow LN2 overclocker, I’m happy to see this. While there are a lot of caveats in some stuff you said, you did an excellent job in packing it all in to a short time-frame. Kudos.
6.9 GHz on 6/9?
Serendipitously nice.
Nice Nice!
As a european *i shall allow it*
@@GameyCat eciN
With 4.20GHz at those e-cores too. What a nice video to watch at 6/9
double nice
Speed of light is also a big limitation, 6.9 ghz means light can only travel 43mm (1.7 inches) per cycle, which is just insane.
The size of the die is 20.5 mm × 10.5 mm, now that clock cycle is only within one core inside that CPU die, but still insane, you'll start to see why there isn't a much of room left.
wAiT a MiNuTe tHoSe SpEcS aRe KiNdA SuS
Also the speed of electricity is slower than the speed of light, which could very well get it to within the range of the die if you tried to get much higher.
@@shadow_rune6178 actually laughed hard out loud.
Splave definitely missed his true calling of being a professional wrestler. I say this as a fan, he's got the PERFECT look for it
He probably does competitive lifting
@@MesquiteAutomotive he does, or at least he did in the past
Gonna be fun next gen with rumours Zen 4 is hitting 5.8Ghz in the lab in early silicon samples, we're gonna have many more attempts at very high clock speeds coming up.
Man that sounds wild. I wonder if they're gonna be hot as fuck or real efficient.
Man that sounds wild. I wonder if they're gonna be hot as fuck or real efficient.
@@batt3ryac1d efficient, zen 4 will be insane
@@batt3ryac1d All CPUs can be both inefficient and efficient, it only depends on what power they run them at from the factory. If you're worried about efficiency just undervolt it like I do.
@@theplayerofus319 shouldn't undervolting not do any harm to it? you're just decreasing the amount of electricity going through the cpu.
Man that really puts it into perspective for how long it's been since I've seen a gpu or cpu actually need to advertise, nature is healing
I was having a heart attack throughout the video, Linus has a reputation for dropping/knocking things over, imaging that happened when he's pouring that liquid nitrogen.
Same here bro 😨😨😨😨
Also.. ever had a LN2 burn? Wear some gloves, not a big hassle but makes such a difference
nah. the leidenfrost effect would save you, so you wouldnt have contact with it, unless you put on gloves and it goes in there
@@bobuccman1424 Nope. I actually was burned by LN2 before, and while the Leidenfrost effect works it doesn't actually protect you from nasty burns :/
Especially with a whole thermos full of it😅
Splave is so chill, probably from being around liquid Nitrogen a lot! 🙃
The guy is so humble. Love the vibe. Thx for the special content. :-)
Linus went from looking like a child, to acting like a child.
I love it
Yeh it was emberassing to watch him.
watching Jay do liquid nitrogen overclocking was really fun so it's good to see you do a video like this! i wouldn't ever try this as i just have no desire to but it is enjoyable to watch
Allen talks and acts like he's the Bob Ross of overclocking and I love it.
2:26 "Take it down to greater than 170 degrees Celsius" Damn... thats hot
You know you're doing some extreme cooling when you need to actually heat it too
Just an FYI (months later) you'll probably encounter the leidenfrost effect with liquid nitrogen. Effectively the liquid boils away and shields the core from cooling; this is why you can dip your hand in liquid nitrogen.
If you convert to a pumped/flow setup you will avoid this. It would also allow you to use other cases like helium with little change.
Source: in grad school we worked with cryogens and temperature controllers for experiments at 4K or below
I love the overclocker duo, that was very enjoyable to watch!!
the new guy looks like a body guard, lol
Do more content with this guy - he is great!
Such a peaceful video. Their voices are soothing
@Linus the true thermal barrier isn't even the increased resistance. It's the reduced impedance due to increased quantum tunnelling as temperature goes up, which increases leakage current, which is the majority source of the increased heat once you get into thermal runaway. That is what will damage the CPU, and that's the reason for throttling. Well... unless you're throttling before you melt the board's power stages or power delivery traces. We're at that point now where some mobos will fail before the CPU!
yeah it sounded wrong to me since semiconductor should have lower resistance as temperature goes up unlike metal. Thanks for clarifying the whole thing.
@@someonestolemyname I was also very confused, as resistance of semiconductors indeed goes down with increasinig temperature.
Ah, this is the post he referenced at the top! Thank you. Now I know that silicon has a negative coefficient of resistivity, which I think explains why he said that the board doesn't like to power on when it's cold- The impedance of the silicon is too high when it's cold & doesn't allow enough current flow.
@@nickpayne1182 At the frequencies they're pushing, 99% of the chip is a series of connected RLC tank resonators being coerced into operating literally in harmony. For best efficiency those little tanks are designed to operate happily in a given frequency range with a given amount of power based on capacitance, inductance, the driving voltage, the driving frequency. The rate of change of voltage is based on the resistance (metal and semiconductor) and impedance (metal and semiconductor.) The resistance of both goes up quite linearly with temperature. Impedance on the other hand is quirky, lol. As the capacitance and inductance deplete they induce current flow which increases impedance to the depletion of the inductance. Higher temps decrease that impedance but prolong it. Net effect is that the transistor takes a little longer to switch, and so requires a higher voltage to switch fast enough to keep up with the clock rate, hence why cooling and under-volting allow you run higher clocks at lower power. Quantum tunnelling does the opposite. It's behaves similarly to how an avalanche starts thanks to vibration. The electrons want to get from A to B. Semiconductors can be biased to operate as either conductors or insulators. Insulation works to prevent electron movement by presenting a uniform wall of electron shells perpendicular to the desired path of the electron(s). As the insulator's molecules heat up and shake larger gaps intermittently appear in that wall. When the gaps get large enough, the odds of electrons jumping through them increases (quantum is all about odds lol). That shakes the molecules even more, and soon you have an avalanche of electrons migrating where they shouldn't, plus more heat than was designed for. If the heat is sufficient it can break molecular bonds and the electrons will start pushing atoms around, aka "electromigration". CPUs have heat sensors to make themselves throttle before that happens, but those sensors can fail and allow a CPU to self-immolate, sadly, and that's part of why TDP exists - the CPU will also throttle above a certain power usage window as a backup failsafe anti-suicide measure. Intel K series processors have that safety feature turned off... you pay for less safety lol.
@@linuxgeex Holy cow, thank you for the explanation & giving me a good starting point to do more research. I learned the basics of electrical theory in my apprenticeship, so luckily this isn't all Greek to me. Your explanation sounds very similar to what I was taught about reactance & hysteresis in AC circuits. Thank you again for the explanation.
I love how Linus literally just made a video to get 6.9GHz on 6/9. Whoever thought of that must have had a NICE day. 😂
It's 9/6 everywhere else in the world. See you on the 6th of September
@@thomasphillips885 it isnt, it's just 9/6 for america, because america wanted to reject everything the english did apparently
Thats nothing don't you know Tom cruise went over 10G!!
It's also at 69k likes rn
Yep.. silly reversed formatting
He is finally overclocking that is awesome
Damn, Splave ole buddy, recognized you right off - looking good Bro! Teammates years ago on the "No Borders Extreme Overclocking Team". He was the best way back then and just kept getting better!
a cool extention of something like this could be to put the whole system in a pressurized nitrogen atmosphere so moisture isn't an issue then use a cryogenic heat pump similar to the ones used to make LN2 as a cpu cooler by maybe using LN2 or helium as the cooling liquid. have a raspberry pi or the like managing flow rates and temps and it could make one hell of a server (wouldn't want a set up that big on a desk so in a rack it must go)
The problem is, the action of the LN2 boiling rapidly is actually what sucks out so much of the heat. It pulls a lot of thermal energy out from the pot to go from the liquid to the gas state. This is actually very much how freon(r134) in our AC systems works, the transition from liquid to gas, and back, makes it able to absorb and release that heat so efficiently
That's a really neat idea tho, I wonder if there is a way to implement it in a moisture free pressure vessel that still allows the energy transfer as efficiently(maybe more?)
Can't they put it in a sealed box with a water condenser .
You can call it a good day when Linus doesn't lose a hand while messing around with liquid nitrogen. I'm pretty sure Yvonne only found out about this AFTER it was already filmed.
it's pretty hard to lose a hand to liquid nitrogen unless you're an absolute fuckup, it vaporizes before touching your skin
@@clavius5734 Gloves are actually bad when handling LN2. Your hands are hot enough when it hits them the LN2 boils and you're protected by the vapour creating an insulation layer (Leidenfrost effect). Put on gloves and the LN2 soaks in leading to v frosty hands. -AC
As was said, you REALLY don't want gloves or even socks when using LN2 for OCing. LN2 isn't really THAT dangerous assuming you know what it does. Movies make people think it'll freeze you instantly and that's just false.
To give a little more context as to the dangers of LN2, it has 2 primarily. First is its displacement of oxygen causing asphyxiation (don't breath the gas). Second is its cold, but when touching your skin it due to the temperature gradient causes the LN2 to flash directly to a gas with a sort of barrier forming between your skin and the LN2. If you put on gloves this doesn't happen, it soaks into the gloves and super cools the gloves leading to burns from it. This will happen with socks and other fabrics too.
I have worked with LN2 for biophysics work. you learn to respect it
Legendary editing at 6:39. God speed friend.
6.9Ghz overCOCKed CPU (Central Penetration Unit), nice specs man. I love your videos
I really love all the additional information, you guys blend in. Good job, this is a 11/10 vid!
@WhatsApp°±I𝟳𝟳𝟱𝟰𝟳𝟲𝟲𝟽𝟪9 get lost scam
This cool guy is like when you go see a car mechanic, big and quiet, but knows their sh*t inside and out,
fantastic vid, one of best you have done lately :D
Alternate title: Linus playing with frozen food and doing a bit overclocking for 17 minutes straight
Alternate alternate title: Linus smashes his banana and pours liquid in the hole
@@charliecharliewhiskey9403 🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣
Or, Linus making a hole in a metal bucket with his banana then pours the liquid in his cpus black hole
15:38 the "wheeeeee" and that look full of concern in the background 🙂
Honestly this is such a cool video. Pushing things to the limits is amazing
Another possibility for potentially avoiding condensation altogether would be to do what you're doing here with the extreme overclocking inside of a walk-in freezer. The cold air tends to be dry and has the bonus of keeping everything cool, but not ridiculously cold. The LN2 would still boil in a walk-in freezer. You just may need to make use of a heat gun or that torch a little more. If you did this inside of a sub-zero environment, you wouldn't need to go to the trouble of building a vacuum enclosure like the one I described in my previous comment. I'm sure you could find someone who would let you borrow their walk-in freezer for a few hours. ...Or you could possibly even buy your own if you wanted. You could maybe use it to continue working towards breaking and even setting your own overclocking records. 😁
Watching him play with the liquid nitrogen had me very concerned until Alex had him put on safety gear xD
Gotta protect the cash cow
Alright! My birthday, June 9th, 1959. Thanks for the birthday gift Linus!
Hot damn those water bottles are amazing. The fact they’re insulating the cold so well is insane
It's vacum insulated, hard to beat.
The 6.9GHz is the first CPU in the market, and, well, it just blows the current offerings away.
You just found a new member of your team. What a cool (overclock) dude.
Loving how the video compression made the black cold look like "Unmoving Plaid" or an end portal.
Awesome! Love the videos partnering with other folks!!!
📩📩☝🏻 ..
..
@7:30 - talking about no money in OC'ing, the "Very True" felt like it came from the gut. I felt that. It was visceral for me lmfao.
Uhhh the new design of the video-references is really good! Love to see it!
You should hire this guy, if he needs a job. I would watch more of him hes chill
Have Splave on more!! Vibing with his teddy bear energy 💪💪 and cool as a cucumber
You guys should do a chilled dielectric liquid cooling, using some of that 3M liquid that’s used for server immersive cooling.
Can't wait to watch this the 6th of september!
I doubt I'll ever overclock anything, but this was fascinating as all hell to watch!
Allen is so cool, the most chilled dude ever compared to Linus. ;)
Splave is the most "chill" guy I have ever seen. Thanks for the video man, these things keep my brain intact :D
Why is chill in quotes?
2:10 Allen in the background just sitting there looking like a super scary bouncer. 🤣
11:59 Wow. Very cool, like this open set up *absolutely* freezing cold, experimenting with these conditions. Awesome stuff O.O
A video about a 6.9GHz CPU released on 6/9/2022, nice.
I'm sorry Linus, but today is 09.06.
06.09 will be in 3 months in September.
Could only imagine having liquid nitrogen running in a full loop watercooled pc
no radiators
ONLY RESEVOIRS
Kingpin already did something like that
Wouldn't it just burst from the LN2 vaporising?
You would need a compressor and a condenser to get the vapour back into a liquid. That might require a stage or too extra (like an LN2-to-ammonia heat exchange and a second compressor loop).
@@xuser48 Apparently, liquid nitrogen is considered impractical as a looped coolant, but CO2 is actually used in commercial refrigeration. Apparently a bit expensive, but I read that it was being marketed as an environmentally friendly alternative to the currently popular refrigerants.
It was interesting to see it suggest big systems might leak hundreds of pounds in a year.
This was the video I needed for my birthday just wish I could have watch on the 9th.
I want to see extreme overclocking being used in a practical setting
Playing fortnite or minecraft
I don't think it have practical use in daily life unless it is used in big tech giants like google to handle their billion search query.
applications horizontally scale too well these days. thats why all the supercomputer's are huge and not just swimming in a vat of LN2.
its not necessary at all this much power. Well i dont know what the hell i need it for anyway lol. Maybe for 2050? who the hell knows.
Watching Splave and Linus together reminds me of the proverb 'Empty vessel makes noise' :|
You guys should try a raised reservoir feeding into the pot governed by a valve that you can modulate the drip rate to keep the rate of cooling/coolant use more accurate.
aah a stable overclocked setup is a blessing!
My kind of CPU.
Nice 😏
Aye my boy. Keep going strong my dude. Your vids helped me with the gtx 970. Been happy with it for half a year now xD
Jack? Is that you?
Jack
hey kryyzp
Only time I've messed with liquid nitrogen was with overclocking a old socket 754 sempron i think 2800+ which was stock like 1.6 ghz pushing it too 4 or 4.1 ghz
On a poor matx gigabyte board with an nforce chipset.
At university ca 2013 or 2014 with a ocz 650w powersupply hooked up too it i think. Agp slot on the motherboard & integrated graphics was first used but that stopped showing a display after a bit (go figure lol.)
So a random agp card was used instead.
Ram was like one or two sticks of 400 mhz cl2 ram at 512 megs or 1 gb in total.
Pointed a fan at them & fan at the vrm which was completely bare btw, no heatsinks at all on either of them.
Classmate got too work building probably one of the most ghetto liquid nitrogen pots ever...
Step 1: Regular size pringles can, don't remember if it was either sourcream or a regular one.
Step 2: Gobs of solder on the bottom of it too raise it up around the motherboard a bit & get a sort of cold surface plate, think it was leaded solder, might have been unleaded or a mix of both.
Ofc making sure not too melt a hole through the pringles can bottom which is kind of aluminium foil ish so close enough right ?
Which also didn't crack from the cold & worked surprisingly well after some sanding too smooth it out flat which took him probably like 35 minutes or so lapping it fairly smooth too the touch & flat. + ofc thermalpaste later used for better contact.
Think it was MX-2 or MX-4 & caked the cpu in it.
Step 3: For mounting the pot somewhat more rigid then gravity, 4x holes where i think wooden sticks or pegs threaded through, zipties through the backside of several mounting spots on the poor motherboard too tighten it down doing it gently a bit at a time too keep it sticking down somewhat squarely on the cpu while making the mounting tight against the cpu.
Ghetto as all hell but worked surprisingly well enough too mess with & get it too boot into the bios at 4 ghz after all, probably not an overclocking record by any means but fun too do anyway.
Also filled the pot with bottlecaps & other random metal garbage from the workshop & a box of random pc screws too absorb & keep the nitrogen from dissipating too quickly, we surprisingly due too the Jank mounting we did had nearly 0 issues with issues booting up having readings at like -102c or so, but might have been the sensor just completely freaking out as well.
Ofc we used paper towels & a bunch of other stuff too insulate the board from any moisture the best we could together with the fans.
If we had a pci ide interface card we probably could have managed too make the poor thing boot into windows but the onboard controller & probably most of the chipset was noping out after like 2.6 ghz already & we kept upping the fsb as long as it booted into the bios after maxing out voltages on chipset, ram & cpu lol.
Ive reached 4.1Ghz or 4.15Ghz too on a sempron; using a phasechange keeping the CPU at -45 on avg. Was a fun day. Installed the phasechange later ontop Ati 2600 pro which had stock clocks of 600Mhz, soldered all caps with larger or better ones, installed a external PSU instead of being fed through the PCI-E Bus, and managed over 1260Mhz core clock lol. Once i turned the voltage way beyond 2.2V the sparks just flew out of the back of the card and the party was over.
I foresee a blooper video: We superchill silicon with liquid nitrogen and, in the interests of science, Linus DROPS it.
Splave looks like he could kill ya with one finger.
Menawhile Splave described in best worlds: the worlds' toughest cinamon roll.
How convenient, 6.9GHz is probably the best you can get. Nice.
And he makes a video on 6/9/22
69GHz
@@rawr51919 in a few years 😳
nah bro... 420GHz on a 6.9 nanometer process
And for the rest of the world, we require a 9.6ghz overclock
You mean 2022.6 GHz? ;-)
And also (and this is maybe the 5th time I write this) you CAN (and should, it's not difficult) put your motheboard in a controlled atmosphere (i. e. a box with a Argon or Nitrogen gas flow) so there is NO MOISTURE in the air. In the research world this is done for example to protect samples exposed to radiation from oxidation.
It's expensive to run a setup like that.
i love how splave's face glows red like he's a little kid going over to his new neighbor's house and seeing that his neighbor also also likes playing with g.i. joe action figures. its so innocent and wholesome 😊🤗
Absolutely, positively insane! Could you somehow run the pc in a vacuum removing air to condensate? Cool stuff guys!
I would think the process of removing air from the vaccum chamber would damage components
If you remove the air and turn it into a vacuum like out of space, wouldn’t the metal components weld to each other and the liquid turn to jelly and evaporated?.
I would guess the latent heat energy absorbed when the LN2 boils away is important for keeping everything cold enough, and a vacuum makes the boiling point too low
@@anthonykoller4459 This only happens with identical metals, and very clean and smooth ones at that
They could, but the issue would be removing heat fast enough, considering it's a vacuum.
You know what else helps? A well sealed room and a dehumidifier to pull more of the moisture out of the air.
Won't get it all but if there are any holes in your moisture barrier it will reduce the risk of anything getting too wet
and then wrap urself in anti static wraps
Insane but I want more videos like this, XD. Try overclock GPU now.
15:11 I like how Alex and Splave are talking about overclocking while Linus there in the middle having fun with the toys.
15:38 Wheee! Whooo!
LN2 is pretty damn cool (pardon the pun) and it looks pretty fun to overclock a CPU with it. Liquid helium on the other hand, although it's even colder than LN2, does annoy me when people waste it on just overclocking a CPU... Helium is a rare resource and is something we're quickly running out of. The more important uses are for cooling magnets in high power particle accelerators and MRI machines - i.e. stuff that really matters.
:( but I really like watching balloons go weeee up up and away. Lol also makes u talk like your talking into a fan now thst is what I would call important
i thought helium was the second most abundant element in the universe
@@anthonycote2916 IIRC It is quite abundant on a scale beyond our planet, but being a extremely light gas that doesn't form heavier compounds like hydrogen does to form water it gets stripped away from the planets atmosphere from radiation. I only did a quick search so this may be wrong but helium and hydrogen have similar concentrations in our atmosphere, but since hydrogen gets trapped as water we have a much larger supply of it.
Also, I remember at school we had some scientists form a lab came in and show stuff with nitrogen. I remember they froze waffles and chocolate for a second and then gave us to try it.
Linus: You know who never fails? Our sponsor MSI!
Me remembering MSIs PC prebuilt: "That's a lie lmao"
It's also a Ventus so it'll probably fail to cool the GPU properly...
Every MSI PC component I ever owned failed in some way within 3 years
@@Eagle3302PL I know someone with an msi laptop, it Shuts down if you rest your palm on the right hand palm rest
I expected Linus to say the guy is world record holder in bench press or something.
0:09 is a very, VERY suspicious "tool" 🤔.
Hmmm, how cool would a Novec submerged PC with a LN2 powered heat exchanger to pull heat out of the Novec be?
I think Novec 7000 boils at 34° C but freezes as low as -122° C.
Submerging the test bench in Novec would remove any condensation risk. It’d also likely stay colder way longer since Leiden frost wouldn’t be an issue, and a few liters of Novec at -100° C would be a massive thermal sink.
I really wish it was possible to buy novec.
@@KeinNiemand same here!
It would be nice to break the 10Ghz mark.
1:33 - we used that trick to quickly "sweep" the lab floor! If you pour nitrogen onto the floor, it propagates in a "wave", carrying dirt and litter until it hits the walls. Then you can just sweep it. Kinda like vacuuming, but much more fun.