honestly it would be much better heat sing for passive aplications heated up air will travel up sucking air trough opening on the bottom because there is an expansion chamber on top heated air will decrease its pressure allowing for more air to be sucked
That’s why tech companies should be led by engineers and not business, legal or marketing people: to enable innovation and advancement to lead the way and speak for itself. Engineers walk the talk.
But in the design process you should also take into account how easily it can be cleaned. I am a computer enthusiast and I disassemble computers on a regular basis. I can say that most of the designs I see are dust traps that in a short period of use will have the very opposite effect. The heatsink will be clogged with dust and the temperature will start to rise. And there will be no easy way to completely clean it. Actually, it will be impossible to clean it, except with the use of an ultrasonic bath cleaner that not everybody owns. And let me tell you that if it is not thoroughly cleaned to be like new, the dust that remains acts like a glue for other dust particles to re-stick on it and accumulate quicker than the first time. In recent years I have noticed that designers do not take into consideration all the parameters of real-world use but also the effect of time itself in the designs of their products. For example, they design a bathroom faucet that may be beautiful in terms of elegance, but in terms of use it is not at all practical because it is not high enough in the sink and to wash your hands you have to touch the walls of the sink that may not be clean. So, you intend to wash your hands and instead you get them dirty (opposite outcome...).
That's a very interesting design, but it is incredibly specialized for its application at its size. I'm curious how well it works in real world performance, like for example on a Ryzen 9 series processor, something that would benefit incredibly from an extremely efficient compact cooler. Dissipating 105 watts of TDP in that small of a package would be no joke.
What exactly does this heatsink optimize? Is it optimized for getting maximum cooling for an arbitrary airflow? Would a different profile work better if you tried to maximize cooling for the amount of noise generated? I would imagine that minimizing noise is more difficult since both the fan and the heatsink profile might contribute to both the cooling performance and the amount of noise generated.
Arbitrary? He says it in the video. A fan is placed on top to blow air down into the heatsink. And of course, a different profile would work better if you had to add noise as a constraint as well. Noise wouldn't be more difficult necessarily. Obviously, they are just having a program run simulations of theoretical designs so it would just apply the exact same principles.
The indian guy has a even thicker unibrow that Mehdi (electroboom). Mehdi should be ashamed :P Their heatsink looks awesome and it apparently does a great job too. I hope these kind of heatsinks will be available for consumers soon.
wondering if anyone has made a heatsink out of silver. I dont think copper and other cheaper metals can transfer heat better than silver. would really help with heat transfer
Large surface area, high thermal conductivity, cheap as peanuts. Copper wire? 300 meters thin copper wire for 100W CPU? Will be more than enough to cool it passively. Mechanical engineering as your guest.
Yes, I do not doubt amazing properties of Copper, but at the speed industry goes, and everything producing more and more heat, I have no doubt that one day copper will be overwhelmed. Maybe Liquid nitrogen is the future...
I think this was just a prototype. We have already some 3d metal printing abilities, so I thibk it should be possible, and obviously it would be better.
I'm not an engineer but I can tell you that 'keeping the air in the heatsink to extract more heat' is BS. He's saying that coz he doesn't know why the simulation produced the curve. Heat transfer works better the higher the temp difference.
Why add unneeded static pressure channels, eddy currents, and most of all why add a fan blowing from top to bottom when heat naturally rises. Highly inefficient looking design. Thats pretty much all thermal mass anyways.
Blowing from top to bottom would be counter flow and is very efficient; you can find natural examples of counter flow in evolved systems such as animal kidneys (the nephron) or fish gills.
I would sincerely appreciate it if you would make me a VRM heatsink for a 2080 ti. I plan to mount an NH-D15 to the GPU, but I want a cleaner heatsink than what’s available online. I’m willing to pay.
I don't think there's a premade VRM heatsink available for the RTX 2080 Ti yet and I don't think Purdue University is gonna make you one. looks like you're going to have to DIY. You might want to check these out for VRM's though modmymods.com/alphacool-gpu-ram-copper-heatsinks-10x10mm-10pcs-17426.html I use similar heatsinks on my GPU and get great performance
i donno... that does NOT look any better than my noctua d15 or even cheap intel cooler. first it has such little surface area and such restricted airflow. where are the test results?
yes, there already exist 3D metal printers. markforged.com/ www.materialise.com/ layertec.be/ (the actual machines & training. They can make the metal denser than you would get by forging even!)
@@dogrudiyosun Then you'll love 3D fabric printing. It's not cloth or wool ofcourse. But extremely flexible material and/or fiber-like material such as acrylic and silk. Kinda like carbon-fiber printing with the metal printers. Still very very early development though. But's it's gonna be here sooner than we might expect. Just like most people never even heard of 3D metal printing while such machines are already in operation.
If you really want to make an efficient heatsink then you would need to use fluid dynamics and AI. Humans can only do so much, and we're not very efficient at calculating fluid dynamics or all the variables that affect it.
Does it work the other way? If it was being actively chilled, could you blow air through it and expect it to pull enough heat from the air even in the short time the air spends in contact with it to be an effective small personal cooler?
Serdar Ozguc's smile at 02:07 is contagious. Congratulations to both of you gentlemen and a job well done!
Finally some good heatsink to my overclocked Core i9-10900k
WOW, FLEX. You are not cool by just flexing your kidney
Same dude nothing can cool my i9 9900k @ 5ghz all cores 😬
@@monkeh3612 nah bro I have a 10700k
Just joking ya know cuz they are hot lmao
@@Sh1denK4i Yeah obviously, try competing with my triple 3090 Founder's Edition, Threadripper 3990X 64-core, 128-threaded, and 2TB of RAM you peasant.
@@Tayfaan Dual GTX295’s on a skulltrail with 5 raptors in raid. Those old builds would chew up power supplies and electric bills like nothing.
The one that doesn't clog with dust always wins.
Just imagine how hard will de dusting off the presented design...
@@Stszelec01 It just needs an integrated port to attach a high-pressure air hose or can of compressed air.
@@rupert274 Yea but it woud require higger pressure than normal canned air
honestly it would be much better heat sing for passive aplications
heated up air will travel up sucking air trough opening on the bottom
because there is an expansion chamber on top heated air will decrease its pressure allowing for more air to be sucked
Part of engineering is design for manufacturing. This design is great for one off designs, but is more complex. See heatsink skiving
That’s why tech companies should be led by engineers and not business, legal or marketing people: to enable innovation and advancement to lead the way and speak for itself. Engineers walk the talk.
Most engineers drink glue for a living
@@fiddlesticks7245 ? I don’t get it
@@fiddlesticks7245 Labelling their drinks as glue is a tactic to reduce the chances of people stealing or tampering with their beverages.
Totally agree. I just remembered Elon Musk's MBA-isation statement.
Ideas don't make functioning companies.
What about adding Diamond Powder/Dust into the mix? How would that change the thermal conductivity of Heatsinks?
But in the design process you should also take into account how easily it can be cleaned. I am a computer enthusiast and I disassemble computers on a regular basis. I can say that most of the designs I see are dust traps that in a short period of use will have the very opposite effect. The heatsink will be clogged with dust and the temperature will start to rise. And there will be no easy way to completely clean it. Actually, it will be impossible to clean it, except with the use of an ultrasonic bath cleaner that not everybody owns. And let me tell you that if it is not thoroughly cleaned to be like new, the dust that remains acts like a glue for other dust particles to re-stick on it and accumulate quicker than the first time. In recent years I have noticed that designers do not take into consideration all the parameters of real-world use but also the effect of time itself in the designs of their products. For example, they design a bathroom faucet that may be beautiful in terms of elegance, but in terms of use it is not at all practical because it is not high enough in the sink and to wash your hands you have to touch the walls of the sink that may not be clean. So, you intend to wash your hands and instead you get them dirty (opposite outcome...).
I noticed that too. This kind of heatsink would benefit from drawing on pre-filtered air.
@@bradykirk9932 You're right.
Compressed air don't care
Do you have a parametric model of this that can be scaled for various applications and form factors?
high there intel research guy :P
is there an absent airflow design version of a 3d printed heatsink but can still spread and dissipate head effieciently?
any update on what happened to this design? is it commercialised yet?
You guys are awesome. I cant wait to use a computer with your heat sink.
That's a very interesting design, but it is incredibly specialized for its application at its size. I'm curious how well it works in real world performance, like for example on a Ryzen 9 series processor, something that would benefit incredibly from an extremely efficient compact cooler. Dissipating 105 watts of TDP in that small of a package would be no joke.
@ungratefulmetalpansy Hail, Grammer Nazi.
It doesn't work in the real world
@@LaithSmadi1 why ?
@@leoyru.3361 I dont really see them making that kind of intricate mesh with metal.
and btw even if they could do that with some fancy 3d printing it wont be economically viable
I bet that would make an awesome water block.
1:22 on the walls put heatpipes.
Bassicaly more surface and air flow better is the heatsink.
I think GYROID heatsink is better. have more surface.
Please show it working!
What do you expect to see? A heat sink literally sit in one spot forever. It doesn't move or does any cool action, what do you expect to see?
@@luongmaihunggia temps on the display reduced
If it actually works
What boundary conditions were used in the competition? Any response would be greatly appreciated
What exactly does this heatsink optimize? Is it optimized for getting maximum cooling for an arbitrary airflow? Would a different profile work better if you tried to maximize cooling for the amount of noise generated? I would imagine that minimizing noise is more difficult since both the fan and the heatsink profile might contribute to both the cooling performance and the amount of noise generated.
Arbitrary? He says it in the video. A fan is placed on top to blow air down into the heatsink. And of course, a different profile would work better if you had to add noise as a constraint as well. Noise wouldn't be more difficult necessarily. Obviously, they are just having a program run simulations of theoretical designs so it would just apply the exact same principles.
STL?
First test of this thing should be done by Linus Tech Tips.
It will definitely give the product a lot of traction and views
electroboom needs to have a chat with this guy about his eyebrows
Excellent work.
Can I invest in this?
What kind of material would be used that can be 3D printed?
Metals can be printed.
everything from thermoplastics to titanium
Maybe just try diamond-heat-sink, it was considered the best material for heat dissipation and that's what we do🤗
The indian guy has a even thicker unibrow that Mehdi (electroboom). Mehdi should be ashamed :P
Their heatsink looks awesome and it apparently does a great job too. I hope these kind of heatsinks will be available for consumers soon.
wondering if anyone has made a heatsink out of silver. I dont think copper and other cheaper metals can transfer heat better than silver. would really help with heat transfer
Large surface area, high thermal conductivity, cheap as peanuts. Copper wire? 300 meters thin copper wire for 100W CPU? Will be more than enough to cool it passively. Mechanical engineering as your guest.
Yes, I do not doubt amazing properties of Copper, but at the speed industry goes, and everything producing more and more heat, I have no doubt that one day copper will be overwhelmed. Maybe Liquid nitrogen is the future...
@@DrZipZwan Cheap graphene may be answer.
@@DrZipZwan Other problem I can see is copper source. It is not renewable resource and will become very expensive at some point.
@@Fosgen True, and even now in some country Copper is already an expensive material compered to its concurrent
@@DrZipZwan At least now, we can enjoy cheap copper. Love that metal, has excellent both, heat and electrical conductivity.
wouldnt it be much smarter to use a more heat conductive material instead?
Like copper or Silver?
I think this was just a prototype. We have already some 3d metal printing abilities, so I thibk it should be possible, and obviously it would be better.
I'm not an engineer but I can tell you that 'keeping the air in the heatsink to extract more heat' is BS. He's saying that coz he doesn't know why the simulation produced the curve. Heat transfer works better the higher the temp difference.
"topology optimization" aka keep pressing buttons until it looks cool.
Nice research.
imagine trying to clean dust out of this
Coolness
Nice video
Okay, so that's not going to fit into a laptop. They need to demonstrate solutions optimized for various common form factors.
This kind of design will still fit say a 1U server which is important for large data centers for instance.
That's really similar to a Brachistochrone curve
Heatpipes still will transport the heat hundreds times faster
helal genç:)
Copper transfer twice than aluminium, big companies can make pure copper heatsink for pennies.
I borrowed the prototype and it melted all over my processor. Now, my fan is stuck to the motherboard.
Did you just put some plastic made prototype over your processor. Omg...
I was thermal modding my thinkpad and somehow I ended here
No one needs an Assault Heatsink
Speak for yourself, my PC is constantly being attacked by Mr Sark, and that sommbitch has been to Nam
Плохой вариант, загрязнение не очистить.
This made of copper mid be the best
Why add unneeded static pressure channels, eddy currents, and most of all why add a fan blowing from top to bottom when heat naturally rises. Highly inefficient looking design. Thats pretty much all thermal mass anyways.
Blowing from top to bottom would be counter flow and is very efficient; you can find natural examples of counter flow in evolved systems such as animal kidneys (the nephron) or fish gills.
I would sincerely appreciate it if you would make me a VRM heatsink for a 2080 ti. I plan to mount an NH-D15 to the GPU, but I want a cleaner heatsink than what’s available online. I’m willing to pay.
I don't think there's a premade VRM heatsink available for the RTX 2080 Ti yet and I don't think Purdue University is gonna make you one. looks like you're going to have to DIY. You might want to check these out for VRM's though modmymods.com/alphacool-gpu-ram-copper-heatsinks-10x10mm-10pcs-17426.html I use similar heatsinks on my GPU and get great performance
bro, takes 2 seconds to shave your monobrow
Welp ........There goes my idea. Its already been done.
Sigh...
Better speak about the construction materials and let aside the shape .............. probably copper is the only efficient one
>Optimizing these heat sinks
>Using aluminum
Ya blew it.
this guys never clean a cpu dissipator.
dissipator?
bet it is 20w tdp lol
pfft, who needs air cooling when you have ln2 and ln2 pots, pfft pfft
Its a turbine, well thought
i donno... that does NOT look any better than my noctua d15 or even cheap intel cooler. first it has such little surface area and such restricted airflow. where are the test results?
congrats you missed the point. that mesh provides a ton of surface area for heat exchange.
Is this even serious?
yes, there already exist 3D metal printers.
markforged.com/
www.materialise.com/
layertec.be/ (the actual machines & training. They can make the metal denser than you would get by forging even!)
@@rakly347 wow bro thanks for the knowledge cheers
@@dogrudiyosun Then you'll love 3D fabric printing. It's not cloth or wool ofcourse. But extremely flexible material and/or fiber-like material such as acrylic and silk. Kinda like carbon-fiber printing with the metal printers.
Still very very early development though. But's it's gonna be here sooner than we might expect. Just like most people never even heard of 3D metal printing while such machines are already in operation.
If you really want to make an efficient heatsink then you would need to use fluid dynamics and AI.
Humans can only do so much, and we're not very efficient at calculating fluid dynamics or all the variables that affect it.
This makes no sense whatsoever.
how compelling, btw how many phds do you have?
@@testify2014 a lot of PHD students in the comments it seems
Does it work the other way? If it was being actively chilled, could you blow air through it and expect it to pull enough heat from the air even in the short time the air spends in contact with it to be an effective small personal cooler?
maybe liquid cooling might be more effective with this. look at the microfin design for waterblocks.
Nice design
Kindly share your email, I am having few queries on the same. I am working on similar aspect in different segments