OCZ's HydroJet Cooler Unveiled At CeBIT 2007

[PV5150]

OCZ has a prototype hybrid cooler that combines air and water cooling techniques to cool down the carbon based core - The HydroJet. More details and pics in the news article.

Link

 

[FarFarAway]

This should be an interesting one for sure

 

[FragTek]

I'm deffo gettin' one if it actually works as planned... I'm a bit leary about the carbon base however, if this is as good as they claim, why has no ones done it?

 

[llwyd]

I'm deffo gettin' one if it actually works as planned... I'm a bit leary about the carbon base however, if this is as good as they claim, why has no ones done it?

Probably the awkwardness of working with it? There are loads of materials that conduct as good / better than copper, nobody ever bothers with them either (probably because a lot of them cost about 5 x more per lb. )

 

[Mr. Smith]

I thought that the model in the picture shown was a non working sample, hence it having a grey plastic insert where the copper base should be?

In the first picture, the base looks to be drawn in copper...

 

[PP Mguire]

If it works better than my Silent Tower (which it looks like it will) im totaly gettin one. It looks neat as hell.

 

[FleX.overclock3d]

Looks very interesting indeed, but very very big!!

 

[JN]

Looks very interesting indeed, but very very big!!

Yeah - I can imagine it needing bolting to the back of the mobo at least

 

[llwyd]

rofl jim that thing is going to need to come with its own scaffolding system

 

[ionicle]

we need a review!

 

[ai_01]

looks like a big watercooled heatsink.

 

[Raging]

looks like a big watercooled heatsink.

it is...

 

[deathwish]

would be interested to find out how this performs

 

[Mr. Smith]

How it works

Tony OCZ

I will explain briefly how this cooler is working.

1 the base is made from directional conductive graphite, we used this material 3 yrs ago but then it was to brittle...now we have a version that is strong enough for direct to CPU contact under pressure.

2 the base has micro channels that over lap each other...this is how we are able to extract so much heat.

3 the fluid is not water, we may use a special coolant but we could use an oil based coolant also.

4 The pump runs off the same impeller as the fan and is centrifugal....it is placed very close to the base and pumps directly on to it, the fluid then travells up the copper tubes to the fluid distribution manifold where it is distributed to the radiators.

5 The radiators have micro channles also, this means they transfer heat at an astounding rate, i have been told the transfer from the base to the radiators of heat under test was measured under a second and the same temp on the base is seen by the radiators...this means this cooler has an ultra high efficiency rating at heat removal.

6 The air cools the radiators on the way into the cooler and then on the way out...total cooling area on the unit here is 0.5M squared...we can make this smaller or bigger depending on how much heat we want to remove.

7 The radiators in the sample here are copper, we have decided to move to Aluminum though to save on weight....there may be a slight drop in thermal performance but its so slight you will never be able to measure it

8 Cooling performance is governed by the temperature of the air entering the unit....deltaT would remain constant for all tests but tested in a warm case will push the temps higher over a test on an open bench.

9 Regarding Cryo Z, units are in production, we are not showing it here but you will see it on sale in the USA soon.

Michael OCZ

So here is a bit more insight: The WayCool technology has been licensed from OnScreen Technologies and we have built more than 50 prototypes over the last year for all kinds of application. This said, copper is reaching its limitations of thermal transfer capability with modern CPUs that run lower power than earlier generations but also have much reduced die area -- therefore higher area power densities than anything before. ... that is up to 200 W/cm2 under normal conditions and way more in overclocked systems.

The only materials that can handle that type of thermals are based on directional graphite, which is the material used in this water block. The design is proprietary and patent pending so I won't comment further.

The second key of the cooler design is the microchannel array contained in each of the heat exchangers. We have achieved designs of less than 1 mm thickness with approximately 80% of the structure being interleaved micro-channels that result in almost complete isothermicity across the entire cooling surface within fractions of a second.

The cooling fluid is moved through a pump - whichever design fits the application, regardless of whether it is a positive displacement, rotary or centrifugal pump or even ion charge pump for smaller units.

The combination of a very high thermal exchange with the heat source with an extremely efficient heat dissipation device that only needs very slow air movement is what this is all about. We built completely passive solutions for graphics cards already that even in a 1 slot form factor beat the crap out of dual slot fan solutions and those will hit the market also in the near future.

The pump and fan are driven by the same motor that can be completely immersed in the coolant - which in that case can also act as noise suppressor for the bearings and as lubricant, similar to dynamic fluid bearings in HDDs.

Since I have spent overall about 3 years in the R&D leading to this prototype, I know very well what it can do and what not and whatever is made public about the technology will hold up...

The biggest problem will mot likely be the IHS because it is not plane in most cases and further, the interface is low temp solder that has a much lower conductivity than anything else in the therml path. We have experimented with fully liquid immersed chips as well but we need cooperation from the chip manufacturers on that front.

That's the main reason why I really don't want to say anything about the final junction temperatures that we can achieve ... it'll depend on the specific chip in question ... not even the model

 

[FarFarAway]

Michael makes a good point about the IHS. Most are badly made cheap POS

 

[pinkpig242]

looks almost like a rad in a block

 

[FragTek]

Intel and AMD both make absolutely horrid IHS's. Intel's are just amazingly craptastic to a whole new level though. You know it's bad when people lap their IHS and actually sand through the corners of the IHS and still haven't started touching the center of the IHS.... Bad Intel! :whack:

 

[PP Mguire]

Now that ive read some more and looked at more pictures i actualy like it even more amazingly. Seems like an AIW watercooling solution that looks like giant air cooled HSF with a different substance on the bottom. Cought my eye and i might get it if it comes out.

 

[Toxcity]

Looks pretty good to me..

 

[deathwish]

hope they make a version for SFF's