Swiftech Withdraws H220 CPU Liquid Cooling Kit from US Market

[newtekie1]

Here's where we're going to have to differentiate opinions on the matter.

Asetek has filed a patent that is basically applicable to any closed loop heat transfer system. You're looking at a patent from a computer hardware focused company, and only seeing that one application.

In reality, houses can use a system of tubes beneath the flooring for heating and cooling. A heat transfer segment, with an integrated pump, can push the heated or cooled liquid out and into the tubes. The tubes then transfer a portion of their heat into the surroundings. These systems have been around in high-end houses and building for decades.

Now, tell me that Asetek has a case. A patentable system must demonstrate that it is a unique solution, and not a trivial combination of existing components.

So, Swiftech is doing the only sensible thing. They are forgoing the huge money pit that this litigation will cost. They are pulling the product that they think might be the source of this litigation, and they are changing their aims.


I cannot call this trolling, because it isn't. I'm calling this a crappy patent, that should never have been approved by the US patent system. Because it would take a law suit to challenge this, Swiftech is backing down. They cannot afford the litigation, but hopefully someone else will.

As far as not suing for other products, just wait. Rarely do companies stop when something is proven viable. If one lawsuit threat stops the AIO competitors, another could well kill the custom loop makers. Who's to say that Asetek won't be designing these systems in the near future? I prefer to allow action to speak, and Asetek's licensing refusal speaks volumes about their intents.



Edit:
Let's clarify one more thing. If you were to spend 20 minutes, and remove all references to CPU and computer what do you have? A patent for a generic closed loop heat transfer system, where the pump happens to be integrated with one of the heat transfer surfaces. I could make the same claims about any car radiator, home heating/cooling system, and the environmental water heating systems used in houses (black sacks on the roof absorb solar energy, heat water, then the heated water is used in the house).

This indicates that the patent was improperly awarded, and any judge presented with it should invalidate it for not meeting basic requirements. If the US patent office was competent and well staffed this crap never would have been given out in the first place.

You went and typed all that out and it was a complete waste. You're still trying to argue based on the assumption they are suing over AOI cooling systems, they aren't. They are suing about the individual, unigue, parts that they developed and patented. I swear I read somewhere that it was the pump/block combo, but BTA said it was the rotating fitting. Either way, they aren't suing simply because they threw a bunch of pre-existing parts together into a system and patented the system. Parts of the system didn't exist before they developed them, that is why they have a valid claim, and these unique parts in the patent are what they are trying to prevent other companies from using.

And if you spend 20 minutes removing CPU and computer, you still have a patent on the pump/block combo and 180° rotating fittings, both of which as far as anyone can tell didn't exists before Asetek developed them. Both of these things are innovative ideas that seem to be unique, no patent judge would invalidate the patent on these.

 

[silkstone]

The patent in question covers that L-shaped fitting on the block/pump, which can turn around 180° while remaining watertight.

"The two patents are intrinsically related: the latter was filed back in 2010, while the former revised the design of the system to better reflect the state of the art when it was filed in 2011. Both include reference to the use of a liquid-cooling system connected to a pump and radiator, designed to be fully integrated into a single maintenance-free design - and if you think that sounds a little broad, you might be right given that the patents encompass 'different embodiments of the heat exchanging system as well as means for establishing and controlling a flow of cooling liquid.'"

and

"The patent in question, publication number US 2012/0061058 (PDF warning), describes 'a cooling system for a computer system [...] comprising a reservoir having an amount of cooling liquid, said cooling liquid intended for accumulating and transferring the thermal energy dissipated from the processing unit to the cooling liquid. The cooling system has a heat exchange interface for providing thermal contact between the processing unit and the cooling liquid for dissipating heat from the processing unit to the cooling liquid.'"

- http://www.bit-tech.net/news/hardware/2013/03/11/asetek-sues-cm/1
- http://www.bit-tech.net/news/hardware/2012/08/31/astek-sues-coolit/1


I haven't read anything about the L-Shaped fitting. But, I have yet to find a well-written (neutral) article on the matter.

 

[lilhasselhoffer]

The patent in question covers that L-shaped fitting on the block/pump, which can turn around 180° while remaining watertight.

What?

Neither patent 362 or 764 make mention of that component. The diagrams provided in the patent filing don't even include provisions for that fitting.

Are you seeing something that I am missing? If so, please point it out so I can understand where you're coming from.

You went and typed all that out and it was a complete waste. You're still trying to argue based on the assumption they are suing over AOI cooling systems, they aren't. They are suing about the individual, unigue, parts that they developed and patented. I swear I read somewhere that it was the pump/block combo, but BTA said it was the rotating fitting. Either way, they aren't suing simply because they threw a bunch of pre-existing parts together into a system and patented the system. Parts of the system didn't exist before they developed them, that is why they have a valid claim, and these unique parts in the patent are what they are trying to prevent other companies from using.

And if you spend 20 minutes removing CPU and computer, you still have a patent on the pump/block combo and 180° rotating fittings, both of which as far as anyone can tell didn't exists before Asetek developed them. Both of these things are innovative ideas that seem to be unique, no patent judge would invalidate the patent on these.

Don't go by what you read, do the reading. Here is the article the other thread linked to, from Swiftech: http://www.swiftech.com/pr-7-19-13-h220-removedfromus.aspx

Their two filings don't make mention of a fitting. What they are (and they're really one filing later updated to reflect the changing reality) is an attempt to make a pump, radiator, block system into a two component system. This is clear as day if you look at the pictures.

My argument is that this is not a novel approach or idea. It's a logical progression. They took a pump, relocated it in the cooling loop, and tried to patent it. The structure of the pump could likely be patented, assuming that the pump worked in some novel way. That is not well defined in the patent.

I am all for patents protecting unique ideas. Asetek likely spent a good chunk of money in designing the system, which was an engineering challenge. What it was not is a novel idea. The purpose of a patent, hopefully we both agree on this, is to protect novel ideas not engineering investments.


As far as the 180 degree bracket, that should be patentable. It is a unique design, that is not a trivial design change. Assuming that is buried in the patent somewhere deep, Asetek has a winner. The problem is that unique idea should have a patent, not the heat change system that is incorported around it.

Neither cited patent from Swiftech covers the bracket, so I don't know how that conclusion has been reached. If there is something other than the press release cited, please do provide it. I'd be happy to be wrong if why I am wrong could be cited, rather than conjectured at.

 

[silkstone]

The pump block combo's original drawing uses Socket 478 mounting, its been in there for a while...

Edit: This is the original patent filed in 2004. It clearly has the pump/block combo in it.

http://www.google.com/patents/EP1923771A1

I just read through this.

It appears to be the original patent + the 2008 update.

Does the unedited 2004 patent contain a water block + pump combination. If not, then it should not even have been considered for a patent due to prior art.

 

[newtekie1]

Don't go by what you read, do the reading. Here is the article the other thread linked to, from Swiftech: http://www.swiftech.com/pr-7-19-13-h220-removedfromus.aspx

Their two filings don't make mention of a fitting. What they are (and they're really one filing later updated to reflect the changing reality) is an attempt to make a pump, radiator, block system into a two component system. This is clear as day if you look at the pictures.

My argument is that this is not a novel approach or idea. It's a logical progression. They took a pump, relocated it in the cooling loop, and tried to patent it. The structure of the pump could likely be patented, assuming that the pump worked in some novel way. That is not well defined in the patent.

I am all for patents protecting unique ideas. Asetek likely spent a good chunk of money in designing the system, which was an engineering challenge. What it was not is a novel idea. The purpose of a patent, hopefully we both agree on this, is to protect novel ideas not engineering investments.


As far as the 180 degree bracket, that should be patentable. It is a unique design, that is not a trivial design change. Assuming that is buried in the patent somewhere deep, Asetek has a winner. The problem is that unique idea should have a patent, not the heat change system that is incorported around it.

Neither cited patent from Swiftech covers the bracket, so I don't know how that conclusion has been reached. If there is something other than the press release cited, please do provide it. I'd be happy to be wrong if why I am wrong could be cited, rather than conjectured at.

You can't use the logic that someone would have thought to do it eventually to invalidate a patent. That is true of most patents. The fact is no one had thought of combining the pump and block before Asetek, it was a unique idea, and patenting it is a valid thing to do.

Tons of patents are just someone taking something and improving it in a new way that no one else thought of, that is the definition of innovation.

I just read through this.

It appears to be the original patent + the 2008 update.

Does the unedited 2004 patent contain a water block + pump combination. If not, then it should not even have been considered for a patent due to it being existing technology.

You can read the original on the USPTO website here. I can't get the image to work, but the descriptions of the figures still match, so it appears that yes the pump/block combo was in the original 2004 filing.

Again, the original art shows 478 mounting hardware, that should give you an idea of how old these drawing are, 775 replaced 478 in 2004 and 478 was pretty much off the market by 2005.

 

[silkstone]

As the patent also mentions a separate res, block, pump and radiator system, I am dubious as to the scrutiny the patent was given. It seems like a patent for a 'pre-assembled' water cooling solution.

We'll see what a judge says at the outcome though. It's a lot of text to read for a Sunday afternoon so I will reserve judgement until a judge rules on it or I see a proper write up on the issue.

Edit - I got the image from the original it appears it was issued in 2005.

 

[DannibusX]

Only read the first page here, so here I go.

Asetek is clearly in the right in demanding licensing fees for their intellectual property. Asetek clearly doesn't meet the requirement of being a patent troll. They actually manufacture AIO units. They invented the pump/block, they patented it at the very least, so it's their property.

Swiftech, etc. should immediately begin the process of license fee negotiations. If Asetek won't license it, then they are dicks, lol.

 

[Mindweaver]

Wow.. Just when I was going to pull the trigger on one of these units..

 

[lilhasselhoffer]

You can't use the logic that someone would have thought to do it eventually to invalidate a patent. That is true of most patents. The fact is no one had thought of combining the pump and block before Asetek, it was a unique idea, and patenting it is a valid thing to do.

Tons of patents are just someone taking something and improving it in a new way that no one else thought of, that is the definition of innovation.

Answer me this question, then either agree to the absurdity or continue with an argument that is untenable.

Tomorrow I am going to file a patent. This patent will be based off of an initially three component system; a pump, a radiator, and a heat transfer surface. Instead of three unique components, I will combine all three into one. It will have the heat transfer surface on the bottom, the pump integrated above that, a flow through a finned radiator, then a series of pipes taking the cooled fluid back down into the heat transfer surface.

The unit itself will look very similar to a conventional large air cooler, with an extra large base.

I am taking three things that were once separate, and combining them into a single device. By your logic, this is patentable. On top of the patent issued to me, I can sue pretty much everyone. I will argue that the competitive options may use a large loop, but selling all three things together violates my patent.


Please, argue that what I am saying is wrong. Argue that this is somehow a derivative work, and it would never get a patent. My argument is that Asetek already has a similar patent. That is the problem I've got. They got a patent for something that should never have been issued a patent. The connectors can reasonably be patented. The pumps, depending upon operational methodology, can be patented. Bolting two existing components together is not novel or unique, and thus should not have been patentable.

 

[newtekie1]

As the patent also mentions a separate res, block, pump and radiator system, I am dubious as to the scrutiny the patent was given. It seems like a patent for a 'pre-assembled' water cooling solution.

We'll see what a judge says at the outcome though. It's a lot of text to read for a Sunday afternoon so I will reserve judgement until a judge rules on it or I see a proper write up on the issue.

Edit - I got the image from the original it appears it was issued in 2005.

http://img.techpowerup.org/130721/Untitled.png

Patents can cover a lot of technology, they do not have to focus on one item. You do not have to patent every individual part. If you patent a system that contains unique parts, then the unique parts are patented as well.

Answer me this question, then either agree to the absurdity or continue with an argument that is untenable.

Tomorrow I am going to file a patent. This patent will be based off of an initially three component system; a pump, a radiator, and a heat transfer surface. Instead of three unique components, I will combine all three into one. It will have the heat transfer surface on the bottom, the pump integrated above that, a flow through a finned radiator, then a series of pipes taking the cooled fluid back down into the heat transfer surface.

The unit itself will look very similar to a conventional large air cooler, with an extra large base.

I am taking three things that were once separate, and combining them into a single device. By your logic, this is patentable. On top of the patent issued to me, I can sue pretty much everyone. I will argue that the competitive options may use a large loop, but selling all three things together violates my patent.


Please, argue that what I am saying is wrong. Argue that this is somehow a derivative work, and it would never get a patent. My argument is that Asetek already has a similar patent. That is the problem I've got. They got a patent for something that should never have been issued a patent. The connectors can reasonably be patented. The pumps, depending upon operational methodology, can be patented. Bolting two existing components together is not novel or unique, and thus should not have been patentable.

Sure, its patentable, but you couldn't sue everyone. They already have prior art. Your concept would improve upon it, compacting the design, and would likely require some pretty nifty engineering to make into on unit that can hang off a CPU socket. You would get the patent, and that engineering that you did would be protected, but you couldn't sue everyone that makes AIO coolers.

You want a perfect example of how the patent system allows people to take something an innovate it just look at barbed wire. There are literally thousands of patents for barbed wire were someone took the original design and just tweaked it. Moved the barbs closer, changed the barb design, changed the material it is made from. Barbed wire was original made of wire, someone decided to make it out of stamped razor thin metal and patented that. Then someone came along and figured out if you run a reinforcing wire down the middle of the stamped metal it worked better, he combined the two previous concept into one and patented that. It was still a patent for barbed wire, but the idea of combining the wire and stamped metal was something no one else had thought to do, and hence was patentable. However, he couldn't go back and sue everyone else that was making barbed wire out of just wire or just stamped metal.

That is why Asetek isn't suing everyone selling closed water loops. If the pump is separate or contained in the res or radiator, they aren't suing them, because they know they really can't.

Also, you'd think combing different parts of a system would be logical and easy to do. But, ironically, Swiftech tried combining the pump with the radiator, they even patented it so no one else could do it. It doesn't work nearly as well as combining the pump with the waterblock. Thee reason is that a lot of people put the radiator at the top of the case, the highest point in the case. So when the fluid gets a little low the pump begins sucking air in this situation. However, putting the pump on the waterblock greatly reduces the likelihood of the pump sucking air.

 

[lilhasselhoffer]

This argument is getting us nowhere. I concede that you have some reasonable points, but the truth is both of our ideas mean little. Going to court is the only way this is going to be solved, and lawyers don't run on logic.


With that, I concede. Continued point and counterpoint doesn't serve either of us.

 

[cdawall]

Answer me this question, then either agree to the absurdity or continue with an argument that is untenable.

Tomorrow I am going to file a patent. This patent will be based off of an initially three component system; a pump, a radiator, and a heat transfer surface. Instead of three unique components, I will combine all three into one. It will have the heat transfer surface on the bottom, the pump integrated above that, a flow through a finned radiator, then a series of pipes taking the cooled fluid back down into the heat transfer surface.

The unit itself will look very similar to a conventional large air cooler, with an extra large base.

I am taking three things that were once separate, and combining them into a single device. By your logic, this is patentable. On top of the patent issued to me, I can sue pretty much everyone. I will argue that the competitive options may use a large loop, but selling all three things together violates my patent.


Please, argue that what I am saying is wrong. Argue that this is somehow a derivative work, and it would never get a patent. My argument is that Asetek already has a similar patent. That is the problem I've got. They got a patent for something that should never have been issued a patent. The connectors can reasonably be patented. The pumps, depending upon operational methodology, can be patented. Bolting two existing components together is not novel or unique, and thus should not have been patentable.

Ultra sold that 5-6 years ago and its already been patented.

 

[jihadjoe]

Bolting two existing components together is not novel or unique, and thus should not have been patentable.

Just about every complex machine is about bolting two or more existing, simpler components together. If we can't patent based on that logic, then by all rights we shouldn't be able to patent anything at all.

 

[bmaverick]

If Asetek even wins against any of the others, it will fail due to the prior Apple G5 LCS that uses the Panasonic pump+block combo. The time Asetek was even fileing for the patents, Panasonic was already in production.

 

[newtekie1]

If Asetek even wins against any of the others, it will fail due to the prior Apple G5 LCS that uses the Panasonic pump+block combo. The time Asetek was even fileing for the patents, Panasonic was already in production.

The Panasonic system didn't use a pump/block combo, the pump was separate from the blocks.

 

[silkstone]

It looks like the pump and block are together. Along with the radiator, that is.x

 

[Grnfinger]

looks very heavy

 

[Norton]

It looks like the pump and block are together. Along with the radiator, that is.x

http://img.techpowerup.org/130722/kuehlblock.jpg

Here's another one:

XIGMATEK AIO-S80DP

*EDIT- Here's an Easter Egg- click Xigmatek's product page for this
http://www.xigmatek.com/product/liquid-aios80dp.php

 

[newtekie1]

It looks like the pump and block are together. Along with the radiator, that is.x

http://img.techpowerup.org/130722/kuehlblock.jpg

I owned one of these, look at it from a different angle, the pump is clearly completely separate from the waterblocks. In fact, anyone into watercooling will recognize the pump they used.

Here's another one:

http://img.techpowerup.org/130722/Xig AIO.jpg
XIGMATEK AIO-S80DP

*EDIT- Here's an Easter Egg- click Xigmatek's product page for this
http://www.xigmatek.com/product/liquid-aios80dp.php

True, but that came out in 2007, long after Asetek's patent, and the pump was still technically a separate unit from the block.

 

[cdawall]

Its a ddc1t with the noise isolator block at the bottom. I had one myself.

 

[Deleted member 24505]

I'd like to find one of them in a market for a tenner

 

[silkstone]

I owned one of these, look at it from a different angle, the pump is clearly completely separate from the waterblocks. In fact, anyone into watercooling will recognize the pump they used.

http://www.the620guy.com/Images/ebay/2011-08-15/Apple G5 Dual 2-7GHz 512K 5.JPG



True, but that came out in 2007, long after Asetek's patent, and the pump was still technically a separate unit from the block.

Couldn't it be argues that as they are bolted together, it is essentially a single unit rather than separate pieces? The tech in the swiftech units does seem different as in the pump has been opened at the bottom and had the waterblock bolted on, but is it a big enough difference?

The meat of their patent seems to cover any type of integrated heatsink, block (+rad), can certain aspects of a patent be invalidated whilst keeping the individual components intact?

 

[johnspack]

Having a hard time getting this... aren't all closed loop cooling systems pretty much the same darn thing? Weird.....

 

[Sp33d Junki3]

Having a hard time getting this... aren't all closed loop cooling systems pretty much the same darn thing? Weird.....

Yes they are, Swiftech H220 and CM Eisberg is not a closed loop. They are watercooling kits, that is pre-built for your convenience.

This is basically gone too far in my books by Asetek. I see the reason to sue Coolit, which is a copy of theirs.
But Swiftech I do not see it, when they been doing this for so long.

 

[newtekie1]

Couldn't it be argues that as they are bolted together, it is essentially a single unit rather than separate pieces? The tech in the swiftech units does seem different as in the pump has been opened at the bottom and had the waterblock bolted on, but is it a big enough difference?

The meat of their patent seems to cover any type of integrated heatsink, block (+rad), can certain aspects of a patent be invalidated whilst keeping the individual components intact?

Being bolted together into a single functioning system is different then making them one solid unit. If you can pull a piece out and the piece still works as it is intended, then it is still technically a separate part. In Asetek's case, the waterblock becomes part of the pump and the pump becomes part of the waterblock, the two will not function separately.

And yes, patent judges can pick apart patents invalidating some parts while upholding others.

Having a hard time getting this... aren't all closed loop cooling systems pretty much the same darn thing? Weird.....

Yes they are, Swiftech H220 and CM Eisberg is not a closed loop. They are watercooling kits, that is pre-built for your convenience.

This is basically gone too far in my books by Asetek. I see the reason to sue Coolit, which is a copy of theirs.
But Swiftech I do not see it, when they been doing this for so long.

This isn't about closed loop coolers, it is about the pump/block combo(and maybe the special fittings).