Asetek is Granted a Liquid Cooling Patent for GPUs

Mmm . . . . there is a little something called prior art which means that unless they are doing something that no one else has done before patent enforcement will be difficult. It would be like trying to patent the wheel.

Yea I agree and thought the exact same thing about prior art. But to me (this is total speculation) from looking at the picture, it looks like they made something like a full cover copper block (or whatever material) that has a heatpipe that lays inside a groove that is channeling heat away from the VRM section towards the middle of the block. This cover also goes to the other side and covers the second VRM and RAM but doesn't have a heat pipe.. just a larger mass of the block over it. Then it looks like they have some sort of square water block that is lays down in the middle over the cover and the heatpipe, and has the water channel. So it's almost like they are using the cover as a large passive cooler, then that small middle square water block to cool the cover and the heatpipe...? Maybe? So possibly since it does not have the water channel travelling over the GPU, VRM, RAM like a full cover block, plus the combination of adding in a heatpipe and what appears to be some sort of modular design, all make it different enough to have a patent granted? I dunno.... I was not a fan of how Asetek ran out Swiftech with previous patents so I'm also a bit leary hearing about this one.

EDIT: Okay just found a copy of the patent and the diagram of the device
http://www.google.com.ar/patents/US8274787

So concerning the picture that is posted in this article with the labeled numbers:

"The embodiment shown in FIG. 4 illustrates an optional arrangement wherein only a single lower planar body 202 is utilized adjacent a single circuit board 10 and associated heat sources, without an upper planar body 204. The heat pipe 208 is routed through a serpentine path to facilitate transfer of thermal energy from the lower planar body 202 to the thermal transfer region 212, defined by a recessed seat or socket for receiving a modular cold plate assembly 100. It should be noted that a recess is not a requirement, and in some embodiments, the cold plate assembly may be removably attached to a thermal interposer assembly component (such as, for example, lower planar body 202 of FIG. 4) by another method. For example, the cold plate assembly 100 may be attached using screws or other fastening mechanisms.

The use of the thermal interposer assembly 200 of the present disclosure provides several advantages for the cooling and temperature management of personal computer adapter cards, circuit boards, or other electronic components. By removably coupling the modular cold plate assembly 100 to a thermal interposer assembly 200 which, in turn, is in thermal contact with the various heat sources on an adapter card 10 or circuit board, the design of the liquid cooling system (not shown) may be standardized and simplified. By facilitating a common cold plate assembly 100 to be used with different thermal interposer assemblies 200, modifying the cooling system is made easier. For example, in order to use a cold plate assembly 100 that was used to cool a first computer adapter card to cool a second computer adapter card (having a different configuration), only the thermal interposer assembly 200 (or one or more of the planar bodies of the assembly) needs be modified. This in turn, reduces development and production cycles, as well as production costs. Replacement or addition of adapter cards 10, to a system does not require breech or modification of the liquid coolant circulation pathways. Modular cold plate assembly 100 provides a standardized component which may be disconnected from a thermal interposer assembly 200 and re-installed in a different thermal interposer assembly without removal from the liquid coolant system or breach of the liquid coolant pathways, thereby reducing the risk of coolant loss, leaks, or the introduction of air into the system."

According to a press release issued by Asetek, the company has been granted a patent related to liquid cooling graphics cards. The release didn't give many details, but did tell us that the patent was granted for Asetek's thermal interposer liquid cooling system built for use on GPUs.

You could have, I don't know, googled it.
https://www.google.com/patents/US8755179?dq=8,755,179&hl=en&sa=X&ei=cGGzU6vuGs6jyASo9oFg&ved=0CBwQ6AEwAA
Going by the description this more of a cooling system with general purpose water cooling blocks that can be snapped onto and off of components on a whim rather then the more purpose built blocks we see now . It also includes provisions for transferring heat from VRMs and RAM to the water cooling block using an appropriate set of heatpipes and card cooler. I don't think this is anything to really worry about, someone making a universal cooler is an interesting step forward but we'll have to see.

I think that as opposed to a custom loop water cooling setup there is no competition yet from the AOI units. I spent a good amount of money on my setup but with the components I used it (in my opinion) looks better than and does in fact perform better than my previous H100. The options to customize and expand are not there yet and until AOIs realize that and offer expansion like the H220, but at a better price point. Right now you need radiators for each part, and unless you have them growing on trees inside your case you can say goodbye to room for 3-4 card configs if you have your CPU cooled by an AOI as well.

I will give it a year or two until we are at that point of customization and 360-420mm rads AOIs will not be worth it in my opinion. Have it simple if you need it, expand to make it more complex if needed.