Asetek Unveils Rad Card Industry's First Slot-In PCIe Radiator Card

[ZoneDymo]

In fairness here, Sapphire may have had the idea before Asetek (tho iirc Thermaltake was the oem for sapphires design) however it was more of a failed concept than a functional product. It was rated for 120w and it's radiator was essentially just a regular heatsink with the coolant piped through the heatpipes.

Giving Sapphire credit for the concept is one thing but saying Aseteks design is a ripoff of theirs is like saying the occulus quest is a ripoff of the virtual boy.

Its not about calling anything a "ripoff", the problem is Asetek claiming an "industry's first".

 

[Cranky5150]

That thing looks like a shite hole covered with a dumpster fire TBH.....it's probably not going to cool worth a damn either..

 

[francisw19]

Kind of a different approach, sure. But most blower fans can get noisy and now you got two of them going. I dunno...I think you'd be better off with a conventional open-air type cooler.

 

[CheapMeat]

Probably is an "industry first" if they're talking about enterprise or system integrator level. Don't forget they do legit certified water-cooling for servers too. The stuff other people listed are cool as heck to me but they're aftermarket consumer enthusiasts products. So it's just buzzword saying for who they're aiming for; more like ServeTheHome, not TPU or BOUTIQUE versus us buying from Newegg/Amazon/eBay. Like, honestly, which of you is actually buying the Alienware Aurora R11. There's a ton more legal work that has to go into these things at that level versus the DIY do at your own risk parts most of us go for. If it doesn't work, breaks, etc, they have to deal with DELL, not you; way bigger risk to them at the corporate deal level. Plus they clearly didn't make it for shits and giggles to piss off the TPU forum. Dell/Alienware ASKED them for it, as it says in the post: When Alienware approached us to develop a solution to enable GPU cooling in space constrained cases, we were up for the challenge," said John Hamill, Chief Operating Officer at Asetek. Practically no one on here would even want the case.

 

[gamefoo21]

Its not about calling anything a "ripoff", the problem is Asetek claiming an "industry's first".

Patent lawsuits coming in... 3... 2... :|

Probably is an "industry first" if they're talking about enterprise or system integrator level. Don't forget they do legit certified water-cooling for servers too. The stuff other people listed are cool as heck to me but they're aftermarket consumer enthusiasts products. So it's just buzzword saying for who they're aiming for; more like ServeTheHome, not TPU or BOUTIQUE versus us buying from Newegg/Amazon/eBay. Like, honestly, which of you is actually buying the Alienware Aurora R11. There's a ton more legal work that has to go into these things at that level versus the DIY do at your own risk parts most of us go for. If it doesn't work, breaks, etc, they have to deal with DELL, not you; way bigger risk to them at the corporate deal level. Plus they clearly didn't make it for shits and giggles to piss off the TPU forum. Dell/Alienware ASKED them for it, as it says in the post: When Alienware approached us to develop a solution to enable GPU cooling in space constrained cases, we were up for the challenge," said John Hamill, Chief Operating Officer at Asetek. Practically no one on here would even want the case.

When they sue anyone for daring to bundle a pump and jet plate... Though apparently the USPTO thinks a pump ramming coolant directly through fins is special...

I wouldn't have a problem if Asstek would innovate instead of just charging those fat patent dollars for the old ass and crap design. Innovate once and sue for the next few decades...

The result is there's not many options for a big company in North America. They get one choice and shit performance.

 

[Chrispy_]

That thing looks like a shite hole covered with a dumpster fire TBH.....it's probably not going to cool worth a damn either..

I've spent a lot of time with rack servers and SFF builds - enough to know that blower designs have their advantages.

Trying to cool 250W with a single-fan, dual-slot blower is going to be disappointingly noisy - but the same design can dissipate 125W in almost total silence - so think of this radiator as proving at least an additional 125W TDP without adding any extra noise. The product image seems to show that the GPU board itself has some cooling (there are vents in the IO plate) so the radiator should theoretically add quiet cooling for an additional 125W on top of whatever the GPU can already provide.

 

[CheapMeat]

@gamefoo21 Yeah but what does that patent issue have to do with this? I prefer pump in CPU block too but DeepCool for example found a way around it and so have many others (pump inside radiator). It's not really just one option though since I believe Corsair and others do make their own variants with said patent. But, oh well. But I don't see how it matters to this GPU slot device. And honestly, I figured the TPU crowd wouldn't go the easy AIO route and do custom instead. Maybe I mistook everyone for enthusiasts. Personally if I care enough to do liquid cooling, I'm going with better CPU blocks, standalone pumps, my own rad choices, etc.

 

[gamefoo21]

@gamefoo21 Yeah but what does that patent issue have to do with this? I prefer pump in CPU block too but DeepCool for example found a way around it and so have many others (pump inside radiator). It's not really just one option though since I believe Corsair and others do make their own variants with said patent. But, oh well. But I don't see how it matters to this GPU slot device. And honestly, I figured the TPU crowd wouldn't go the easy AIO route and do custom instead. Maybe I mistook everyone for enthusiasts. Personally if I care enough to do liquid cooling, I'm going with better CPU blocks, standalone pumps, my own rad choices, etc.

Corsair buys their AIOs directly from Asetek slightly modified.

Swiftech had to scrap their original design, and had to make it use DDC pump in the stack. This is because the pump can't be integral to the cold plate or the top. The MCP50X style of pump needs a bearing support built into the top... Suddenly the pump is integral... Asetek said to a judge look at our noodles piece of shit pump that pushes through a crap cold plate design, and won. The result is the Apogee Drive II... Swiftech can't even sell it in a bundle with a radiator and hose... For fear Asetek can call a kit, an AIO. The X2 and X3 are pump on rads and suffer some significant penalties as a result

Ironically in Europe and Asia, you can get many AIOs that are pump on block, that crush Asetek.

As for the patent in question... Cooler Master's design for the Fury X was so much better that AMD paid off Asetek to continue using a superior design. That's something...

There's nothing unique about the Asetek design, and for them to act all special is an act that deserved to be hit by a water bomber. That design is obviously lifted from the Sapphire/ThermalTake cooler and they don't deserve to claim anything as first.

I get why Asetek is stuck, they have no pump IP beyond that thing they've been using for decades that'll fit. The biggest change I've seen is on some Corsair AIOs they have a slightly larger impeller or more blades...

I had to smuggle my original 220 in from Europe and I live in Canada.

I have built many custom loops since then, but AIOs are the gateway, and the patent that Asetek uses to troll and stifle innovation should have died years ago.

 

[TheinsanegamerN]

So they took the loud turbine fans from a blower card, and added a second one.

I dont get it. How is this desirable? The limitations of airflow and the noise of high RPM speeds are still present here, you've just added expensive water-cooling to the flawed blower design. A quality 240mm radiator and 120mm fans will blow this thing out of the water performance wise.

 

[John Naylor]

Crap design, crap materials can only lead to crap performance ... Good science project on glavanic corrosion in cluded for ya kid i'd expect

 

[Valantar]

So they took the loud turbine fans from a blower card, and added a second one.

I dont get it. How is this desirable? The limitations of airflow and the noise of high RPM speeds are still present here, you've just added expensive water-cooling to the flawed blower design. A quality 240mm radiator and 120mm fans will blow this thing out of the water performance wise.

Given the design you're getting a lot of radiator fin area in a small space, cooled by a relatively small fan without the need for a 120mm or 140mm fan opening in the case. Radial fans have ridiculous static pressure, so it won't need to run all that fast to push a decent amount of air through the radiator. And given the thermal mass of water, the improved dissipation from the die compared to an air cooler, and the likely size of the radiator, this will likely be competitive with any 120mm AIO-equipped GPU without the need for Dell to redesign their "small" gaming case and at perfectly reasonable noise levels. There's no reason to expect this to get as loud as a standard blower GPU cooler simply due to the (much) larger fin area and improved thermal transfer compared to the tiny fin stacks on dual slot blower coolers.

 

[TechLurker]

What they should really do is a 3 slot water cooling shroud replacement for the stock cooling with everything integrated into it, no visible pipes, pump, radiators, replacement backplate exhaust for all the airflow.

Reminds me of the ASUS 2080 Ti Matrix. It's effectively an AIO on top of a GPU. Odds are likely that ASUS could pushback on Asetek for copying such an idea, although if one digs into the details, Asetek has a tiny bit of wiggle room in that they can use their crappy pump combo mated to a similar radiator/fan stack. Linus covered the design in some detail, and how ASUS managed to avoid any potential Asetek infringement. That said, it'd be nice if ASUS could work with EKWB to do a GPU AIO kit that emulates the Matrix to some degree; use ASUS' own pump design, and EK's experience to create a 3-4 slot GPU AIO Block and 92mm fans, as well as a smaller 3-4 slot GPU AIO block that uses smaller 40mm fans for half-height chassis.

 

[TheinsanegamerN]

Given the design you're getting a lot of radiator fin area in a small space, cooled by a relatively small fan without the need for a 120mm or 140mm fan opening in the case. Radial fans have ridiculous static pressure, so it won't need to run all that fast to push a decent amount of air through the radiator. And given the thermal mass of water, the improved dissipation from the die compared to an air cooler, and the likely size of the radiator, this will likely be competitive with any 120mm AIO-equipped GPU without the need for Dell to redesign their "small" gaming case and at perfectly reasonable noise levels. There's no reason to expect this to get as loud as a standard blower GPU cooler simply due to the (much) larger fin area and improved thermal transfer compared to the tiny fin stacks on dual slot blower coolers.

So why do blower GPUs run hot and loud if there is so much room and radial fans are so great?Surely the removal of the PCB isnt freeing up so much space the radiator is magically twice the size of a blower GPU heatsink.

 

[CheapMeat]

Have you see the case? They're better in confined spaces and cases where you want the hot air to directly exhaust.

 

[Valantar]

So why do blower GPUs run hot and loud if there is so much room and radial fans are so great?Surely the removal of the PCB isnt freeing up so much space the radiator is magically twice the size of a blower GPU heatsink.

Have you seen a disassembled blower GPU and its fin stack? They are tiny! I fully expect this radiator to be at least twice the volume, if not larger. The absence of a PCB is one thing, the absence of the components on the PCB would be the real space saver - no chunky coils, video outputs or other things that eat into where you can have fins. So yes, I would indeed expect this radiator to have a much larger effective surface area than a standard blower fin stack. Also, radiators have excellent heat transfer across their entire area (most blower GPUs don't even have heat pipes, even if premium versions tend to use vapor chambers) which will definitely improve things. Now, I'm not saying it will be better than something like a 120mm rad with comparable surface area, but the entire point is that this allows for installation in cases where you don't have room for that, and the design seems to make the most of such a use case. As long as they're using a reasonably good blower fan I expect this to cool well and be quite tolerable in terms of noise.

 

[Chrispy_]

Crap design, crap materials can only lead to crap performance ... Good science project on glavanic corrosion in cluded for ya kid i'd expect

i'm not keen on mixed-metal loops for the galvanic corrosion reason, but this product is specific to a GPU which has a useful lifespan of maybe 4 years before it's too slow and outdated to be relevant. In the context of a factory-sealed, corrosion-inhibited coolant with a 4 year useful lifespan, I think we can stop worrying about galvanic corrosion - It won't be a problem for the useful lifespan of the GPU it's attached to.

I still think mixed-metal AIOs for CPUs are a terrible idea because those can be re-used for your next CPU - at least they could be if they weren't corroded to hell and only fit for the trash at that point....

@Valantar also has a good point about the size of this blower that I'd missed - without anything in that card except a radiator it's likely to have twice the cooling surface area and no display outputs blocking the exhaust slots at the end of the fin stack. Given that I can cool my 5700XT almost silently at 125W* (Fan at 20% or just 1150RPM) then I would imagine the increased surface area and exhaust area of this card should be able to deliver 250W cooling at extremely low noise levels - WITHOUT heating up anything else in the case.

That's a win, IMO.


* - In case anyone's interested, I've taken a 6% performance loss in benchmarks for the privilege of silence and run at 1733MHz boost clock, 905mv - tested stable in OCCT for 4 hours at max shader complexity. IF you have an AMD reference card then undervolting is something you should investigate.

 

[svan71]

interesting looking case, I have yet to see a Alinware desktop without a mess of and mis matched wire colors and cables. Always amazes me what people pay for these computers and how little dell cares about what it looks like inside.

 

[BoboOOZ]

@Valantar also has a good point about the size of this blower that I'd missed - without anything in that card except a radiator it's likely to have twice the cooling surface area and no display outputs blocking the exhaust slots at the end of the fin stack. Given that I can cool my 5700XT almost silently at 125W* (Fan at 20% or just 1150RPM) then I would imagine the increased surface area and exhaust area of this card should be able to deliver 250W cooling at extremely low noise levels - WITHOUT heating up anything else in the case.

That's a win, IMO.

Yepp, looks like it has decent potential, we would still need a review to be sure, but I think many users have PCI slots to spare, while most high end GPU's have >200W TDP's that require more cooling estate.

* - In case anyone's interested, I've taken a 6% performance loss in benchmarks for the privilege of silence and run at 1733MHz boost clock, 905mv - tested stable in OCCT for 4 hours at max shader complexity. IF you have an AMD reference card then undervolting is something you should investigate.

Since I'm playing in 1080p, mine is staying at 100W most of the time, with the fans not turning at all. However, when I buy a WQHD monitor, I will probably need some additional modding to retain the same acoustic level.

 

[XL-R8R]

The state of the comments (from some long time users, to boot!) on this forum of recent times is a bit disgusting; if you dont like something, dont waste time commenting or talking utter shit... just dont post.



As for the cooler, I for one will be happy to see some results before ranting/rambling my balls off on some online forum and looking entitled OR deranged in the process.....





Get a grip, whiners.

 

[Chaitanya]

Corsair buys their AIOs directly from Asetek slightly modified.

Swiftech had to scrap their original design, and had to make it use DDC pump in the stack. This is because the pump can't be integral to the cold plate or the top. The MCP50X style of pump needs a bearing support built into the top... Suddenly the pump is integral... Asetek said to a judge look at our noodles piece of shit pump that pushes through a crap cold plate design, and won. The result is the Apogee Drive II... Swiftech can't even sell it in a bundle with a radiator and hose... For fear Asetek can call a kit, an AIO. The X2 and X3 are pump on rads and suffer some significant penalties as a result

Ironically in Europe and Asia, you can get many AIOs that are pump on block, that crush Asetek.

As for the patent in question... Cooler Master's design for the Fury X was so much better that AMD paid off Asetek to continue using a superior design. That's something...

There's nothing unique about the Asetek design, and for them to act all special is an act that deserved to be hit by a water bomber. That design is obviously lifted from the Sapphire/ThermalTake cooler and they don't deserve to claim anything as first.

I get why Asetek is stuck, they have no pump IP beyond that thing they've been using for decades that'll fit. The biggest change I've seen is on some Corsair AIOs they have a slightly larger impeller or more blades...

I had to smuggle my original 220 in from Europe and I live in Canada.

I have built many custom loops since then, but AIOs are the gateway, and the patent that Asetek uses to troll and stifle innovation should have died years ago.

You should read other Asstek Patents as well, they basically can stiffle custom loops as well if they are motivated enough. Also the AIO described in their very 1st patent looks like Cooler Master Aquagate Mini R80/120 series of AIO which came nearly 1-1.5 year before Asstek was awarded a patent.

 

[Berfs1]

The stupid part about it is that it needs a PCIe connector. How about not requiring a PCIe finger when it doesn't take power from PCIe? (if it does, then ignore this) This will allow for being able to fit in the radiator almost anywhere. While I do understand it may need the finger for stability, perhaps not making it require a PCIe x16 slot and making it require x1 with the x16 lock at the end?

 

[ZoneDymo]

interesting looking case, I have yet to see a Alinware desktop without a mess of and mis matched wire colors and cables. Always amazes me what people pay for these computers and how little dell cares about what it looks like inside.

View attachment 155635

ermm its specifically for those people though, those that dont know nor care about the inside, they just want a "badass gaming pc" and that is it.

 

[phanbuey]

Sorry, but this is really crap design. It wastes PCIe slots for a starter, which is a no go.
This is also going to be worse than a regular radiator to get rid of the heat and that blower fan, just urgh...

But it's for OEMs like HP that don't even use that space to begin with. The new omen is a ATX chassis with 1 PCIE slot, and no airflow so this would be perfect.

ermm its specifically for those people though, those that dont know nor care about the inside, they just want a "badass gaming pc" and that is it.

I just did a rebuild for a guy with one of those... I don't know if you guys ever heard or interacted with one of these in real life but they are LOUD. At gaming load that thing sounded like a vacuum.

It also weighs a ton and has nowhere good to pick it up by due to the shape. Kind of a disaster all around. Even if you're just a casual user. The guy was so blown away but how much quieter, lighter, and easier to manage the same exact system with a new cooler in just a normal NZXT case was.

 

[Animalpak]

Here is my AIO mod on a Sapphire Vega 64
View attachment 155302

The RED fits soooo well to ATI ( once ) AMD ( now ) video cards. A color that distinguish them since... The first graphics card they put on the market. And the reason is a well known reason they run HELL RED HOT.

 

[Mayclore]

I can appreciate an interesting concept, but, man, I dunno about the efficacy of something like this.