Water Cooling is DEAD. Meet the THERMOSIPHON!

[Narmonteam]

8 minutes ago, Bitter said:

Because Canada is still British, that's why they're always apologizing.

Ah, I guess the video was filmed on rememberance day, kinda had me confused why he wore it today

[Bitter]

1 minute ago, Juular said:

Nah, by the description on their site it's smth like long-ass vapor-chamber thingy. I was wondering why cooler manufacturers don't make vapor chambers anymore, might be too expensive, looks like they've managed to make it somewhat cheaper and still effective.

It's not a vapor chamber, it's a thermo-siphon which is why it only works in specific orientations. It boils a working fluid and the bubbles push it into a condenser where it cools and then is sucked back down by the pressure differential, the fluid acting as it's own check valve. It's pretty sensitive to changes in angle and what not so the fact they got it to work flat and on it's side isn't bad.

[Juular]

2 minutes ago, Deli said:

I'm afraid Linus has to eat his word when the production model arrives. The cooler may actually be good. But the video title is setting the bar too high claiming watercooling is dead.

If YT reviewers were to eat their words every time their click-bait title were proven wrong or misleading i pretty sure they all would've been long dead by words overdose.

Just now, Bitter said:

It's not a vapor chamber, it's a thermo-siphon which is why it only works in specific orientations. It boils a working fluid and the bubbles push it into a condenser where it cools and then is sucked back down by the pressure differential, the fluid acting as it's own check valve. It's pretty sensitive to changes in angle and what not so the fact they got it to work flat and on it's side isn't bad.

Uh. I'm no expert in hydrodynamics\thermodynamics, what's difference with vapor champer \ heat pipe ? I thought they worked pretty much the same.

[Eigenvektor]

6 minutes ago, Juular said:

Uh. I'm no expert in hydrodynamics\thermodynamics, what's difference with vapor champer \ heat pipe ? I thought they worked pretty much the same.

The technical implementation. Linus actually explains it in the video. But here's a quote from Google:

Quote

In the heat pipes, the working fluid returns from the condenser to the evaporator due to capillary pumping effect while in the thermosyphon the working fluid returns exclusively by gravity because it has no capillary structure.

 

[Juular]

1 minute ago, Eigenvektor said:

The technical implementation. Linus actually explains it in the video. But here's a quote from Google:

I didn't watch 20 minute video on the topic that could've been delivered in a few pages of written article.

So it's advantage is that it has faster coolant movement and supposedly heat transfer than heat pipes but works only by gravity in specific orientation ?

[Vorg]

Still too tall to fit in a 4u case.

[Eigenvektor]

9 minutes ago, Juular said:

I didn't watch 20 minute video on the topic that could've been delivered in a few pages of written article.

So it's advantage is that it has faster coolant movement and supposedly heat transfer than heat pipes but works only by gravity in specific orientation ?

Just saying the info is there Heat pipes can become "stuck" when the heat is too great for the liquid to condense. This greatly reduces heat transfer when you exceed that temperature. A thermosyphon doesn't have that problem, apparently. They usually only work in one orientation, but supposedly this is something Ice Giant was able to solve so their cooler can work in a sideways configuration.

 

Here's the timestamp, the explanation is about 1-2 minutes in length.

 

[Bananasplit_00]

i want this thing, so bad, but the preoders are only for NA markets

 

[RejZoR]

This is not a new tech. AKASA made a thermosyphon cooler years ago with AKASA Revo AK-925. It just didn't perform exceptionally well because of small format and single fan and probably some other factors or manufacturing processes.

 

http://www.frostytech.com/articleview.cfm?articleID=2576

 

Then again neither were heatpipes and they were pretty underwhelming until they started using tower heatpipe design. It's when it really peaked in efficiency.

 

This insane performance is also thanks to massive fins and Delta fans at max speed. If they can slim it down and use quieter fans, I'm interested. Though I'd prefer AiO design so I'd place the heat output directly on the case exhaust. Otherwise you need to double the fans, one for the cooler and one for the case to exhaust the cooler output heat.

[Domrockt]

I would love to see a 9900k @ 5ghz 1.4V avx load test with lets say 24°C Ambient temp.

 

That would make, float my boat.

I even think about a mod with that novec stuff that the 8auer uses.

[orbitalbuzzsaw]

2 hours ago, Den-Fi said:

I'll take 10.

 

Also, lol... this is the first time I've ever seen a coming soon for a review...

Yo, what

[Bitter]

4 minutes ago, RejZoR said:

Though I'd prefer AiO design so I'd place the heat output directly on the case exhaust. Otherwise you need to double the fans, one for the cooler and one for the case to exhaust the cooler output heat.

Sounds like ducting is your friend.

[brandishwar]

You skipped one generation of heatsink. With the Pentium 4, the original heatsinks were pure copper due to how much heat the P4 output. That's why you needed to bolt them through the mainboard into standoffs around the CPU socket. That was also before you had support plates behind the CPU socket as well, given that today's tower coolers are much heavier, but thicker (slightly) mainboards combined with the support plate behind the socket allow that to work.

[Bitter]

26 minutes ago, brandishwar said:

You skipped one generation of heatsink. With the Pentium 4, the original heatsinks were pure copper due to how much heat the P4 output. That's why you needed to bolt them through the mainboard into standoffs around the CPU socket. That was also before you had support plates behind the CPU socket as well, given that today's tower coolers are much heavier, but thicker (slightly) mainboards combined with the support plate behind the socket allow that to work.

Mmmm not all copper, copper slug in the middle though to absorb the sudden jumps in heat and then aluminum fins to disperse it. Aluminum sheds heat faster but copper absorbs it better, this works great for shaking off the heat from a CPU that's spiking power up and down rapidly. The 478 P4's lacked any kind of dynamic frequency scaling but they didn't output 100% heat 100% of the time, and with hyper threading added later boy they could get HOT...PresHOT as a matter of fact lol.

[amdorintel]

1 hour ago, Bitter said:

PresHOT

Biggly HOT!

 

Yeah I remember those times, they got super duper got, big league hot! now where my Covfefe

[sydh]

Where was the temp testing in the second orientation of the cooler? Or was that supposed to be in the release version?

[Luscious]

If they're using Delta fans you can expect the fans to last 10 years as well, or at least 5 if you run them 24/7. They may be loud but they can move air unlike any other. I have Delta fans in my rig and absolutely love them.

[Fierybattlehawk]

This cooler design has huge potential and I'd love to see it maybe even passively cool 95W CPUs (within a reasonably small size) in the near future. It amazes me that this is equivalent if not better than watercooling and CHEAPER than a 280MM or bigger AIO with the MSRP being 120 USD and most (good quality) 280MM+ AIOs are over 140 USD. Also take into account no pump required and thus better reliability.

[Bitter]

3 hours ago, amdorintel said:

Biggly HOT!

 

Yeah I remember those times, they got super duper got, big league hot! now where my Covfefe

Which is funny cause in comparison they weren't that hot it was just at the time they seemed really hot, but now 8 core desktop CPU's are becoming the norm and so are 150W TDP's. That 2700X of mine, I got her down to 190W under small fft Prime95 now with a decent 0.080 undervolt. That's still a fairly big  number though!

[Juular]

2 hours ago, Luscious said:

If they're using Delta fans you can expect the fans to last 10 years as well, or at least 5 if you run them 24/7. They may be loud but they can move air unlike any other. I have Delta fans in my rig and absolutely love them.

No way they would ship the end product with Delta fans, that's prototype, they're like 30-50$ per piece and this thing costs 130$ for preorder. This "review" are completely useless really, it's just a proof of concept, we need to wait for reviews of real product.

[amdorintel]

18 minutes ago, Juular said:

This "review" are completely useless really

thats the way its becoming, oh well clickbait this clickbait that, all for what? clicks! lol

 

18 minutes ago, Juular said:

we need to wait for reviews of real product.

anywhere else but linus technical bulletin board then

[Levent]

FYI: it's Yunus A. Çengel not Yunun A Cengel. He is a pretty notable figure in fluid dynamics so I am not sure how that error slipped.

[Juular]

42 minutes ago, VegetableStu said:

oh heck yes, they posted dimensions

Yeah, that's for end product, with heatsink shrinked to 31mm thick from 103mm thick on the prototype which LMM were reviewing, no way it would perform the same with triple smaller heatsink and regular sub 2k RPM fans.

 

Water cooling are most deifinitely not dead.

[nutral]

I'm really curious on what will happen with the full size model. because size is the one thing a thermosyphon has a good advantage.

Heatpipes are really effective and can transfer a lot of heat, only issue is:

Temperature range based on filling pressure/medium

They get less effective the longer they get

They get less effective the larger the diameter is.

 

Something like a thermosyphon does not have some of these issues and thus is really easy to scale up.

Altough you still get the same issues of having a range based on filling pressure and medium.

The other issue is that you can get entrainment of fluid in the gas.

 

I wonder what kind of medium is used and at what pressure.

What probably happens is that the fluid is first heated up to saturation point, and then starts to evaporate and when more heating is required, the pressure rises, which makes the gas more dense and makes the boiling point of the fluid rise. So it can increase effectiveness when under a higher load.

 

I am curious on what kind of performance we would get on lower heat inputs like a 6 core intel cpu and

if it is possible to blow this cooler up

I believe heatpipes are filled with an amount so that when it is fully evaporated, it would still be below design pressure.

If this device does have a lot more liquid, when you turn off the fan for example, it should keep evaporating until the source temp is lower than the boiling point or until it reach an equilibrium, or until it blows up.

It could also be it is also filled with an certain amount so that it can't go over the design pressure.

 

I wonder if they mention a maximum temperature and/or heatinput in their manual.

[AlexTheGreatish]

2 minutes ago, nutral said:

I am curious on what kind of performance we would get on lower heat inputs like a 6 core intel cpu and

if it is possible to blow this cooler up

I believe heatpipes are filled with an amount so that when it is fully evaporated, it would still be below design pressure.

If this device does have a lot more liquid, when you turn off the fan for example, it should keep evaporating until the source temp is lower than the boiling point or until it reach an equilibrium, or until it blows up.

It could also be it is also filled with an certain amount so that it can't go over the design pressure.

 

I wonder if they mention a maximum temperature and/or heatinput in their manual.

I would really not be concerned about it blowing up under any heat load you could achieve with a computer.  It is possible to blow us a regular heatpipe if you use a torch or something, and you probably could do the dame with the Thmerosiphon - but unless you're putting it in the oven on broil you're probably fine.