gpu \ cpu all in one cooler

So... what, do you have to go DIY on the rest of your graphics card with heatsinks? That GPU cooler looks cheap and flimsy.

If you're on a budget, I'd almost rather go with EK's new all-aluminum kits than this thing.
 
jonw.gif


Outside space constraints--and one would need the room for a 360 rad...
Why?
If one is looking for cool running equipment, it would be better to separate out the components, 240x2 or 120 + 240, and done with AIO liquid cooling components.
 
jonw.gif


Outside space constraints--and one would need the room for a 360 rad...
Why?
If one is looking for cool running equipment, it would be better to separate out the components, 240x2 or 120 + 240, and done with AIO liquid cooling components.
Hmm... Not sure about that. There's a non-trivial amount of testing that claims for a given radiator area, a single loop is more effective than dual loops, put simply because use cases which stress CPU and GPU simultaneously are rare.
 
Agreed, but I would assume that testing was done on custom loops, and not AIO's.
 
Agreed, but I would assume that testing was done on custom loops, and not AIO's.
Sure. Would we expect the results to be different though? The same reasoning applies regardless of whether you're using a custom loop or an AIO: CPU+GPU heat dump situations are rare.
 
I think we need to keep in mind now low the flow rate and head is on CLC pumps .. it's 1/2 or less of what any AIO open loop is and even less when compared to custom loop pumps. I believe their low flow rate combined with aluminum radiators is a problem .. and the resulting noise of the high flow / high power fans needed to supplying the needed airflow to cool those aluminum radiators only adds to their negative attributes. :confused:
 
I think we need to keep in mind now low the flow rate and head is on CLC pumps .. it's 1/2 or less of what any AIO open loop is and even less when compared to custom loop pumps. I believe their low flow rate combined with aluminum radiators is a problem .. and the resulting noise of the high flow / high power fans needed to supplying the needed airflow to cool those aluminum radiators only adds to their negative attributes. :confused:
Well, while the flow rates on these pumps look pretty awful to a person used to custom watercooling, we can't discount the fact that they still work.

I recently chopped up a Corsair H100i GTX AIO and added a reservoir to it to make kindof a faux custom loop. The flow rate on that pump is... it's just awful. When bubbles are rising against the flow in a vertical section of the tubing, the bubbles win. Bleeding is nearly impossible.

All that said, the thing still keeps the GTX 970 it's attached to at 45c max under full load.

This kit has two pumps, one for each block... so there's that to its credit.
 
Well, while the flow rates on these pumps look pretty awful to a person used to custom watercooling, we can't discount the fact that they still work.

I recently chopped up a Corsair H100i GTX AIO and added a reservoir to it to make kindof a faux custom loop. The flow rate on that pump is... it's just awful. When bubbles are rising against the flow in a vertical section of the tubing, the bubbles win. Bleeding is nearly impossible.

All that said, the thing still keeps the GTX 970 it's attached to at 45c max under full load.

This kit has two pumps, one for each block... so there's that to its credit.
I think it's more like 'they manage to flow enough' rather than 'they still work'. :D

Like you said, 'flow rate of that pump is... it's just awful.' Just out of curiosity, can you pee a stronger flow than the H100i pump can? :D

I've been compiling a list of AIO (open loop) and CLC pump flow rate, head, rpm, amp load, dBA and it's pretty scary. Many don't give much if any specs, and some only the amp rating. At least with amp rating we have a pretty good idea of flow rate & head.
https://hardforum.com/threads/might-be-done-with-aios.1937312/page-2#post-1043099563
 
Last edited:
I think it's more like 'they manage to flow enough' rather than 'they still work'. :D

Like you said, 'flow rate of that pump is... it's just awful.' Just out of curiosity, can you pee a stronger flow than the H100i pump can? :D

I've been compiling a list of AIO (open loop) and CLC pump flow rate, head, rpm, amp load, dBA and it's pretty scary. Many don't give much if any specs, and some only the amp rating. At least with amp rating we have a pretty good idea of flow rate & head.
https://hardforum.com/threads/might-be-done-with-aios.1937312/page-2#post-1043099563
"They manage to flow enough" = "they work."

I recognize that they may be beneath the snobbery of us custom watercoolers, but I'm an engineer. Good enough is perfect. I think we tend to make much too much of a such about flow rates around here, to be honest. =p

Head isn't really important with CLCs anyway. They're filled and sealed, so they never have to make that initial rise through the system.
 
Last edited:
"They manage to flow enough" = "they work."

I recognize that they may be beneath the snobbery of us custom watercoolers, but I'm an engineer. Good enough is perfect. I think we tend to make much too much of a such about flow rates around here, to be honest. =p
Head isn't really important with CLCs anyway. They're filled and sealed, so they never have to make that initial rise through the system.
For me it's is nothing to do with 'snobbery' and everything to do with function. I agree many probably think they need way more L/h than they need, Most brand new CLCs are about equal to top air coolers. After a couple years use their performance is below that of top air coolers. A good custom loop is significantly better than top air cooler both when it is new and when it is 10 years old.

As an engineer you should understand the difference between 'just enough' to get by and more than just enough so as heat increases more coolant can flow through waterblock keeping it cooler .. or as pump wears the needed flow can be maintained. Head is important because it translates into pump's ability to overcome friction / resistance within the loop .. and as an engineer you should know liquids flowing through hose, waterblock and radiator cores are fighting friction. ;)

I agree, few of us need the kind of flow rat a D5 series pump is capable of (up to 1500 L/h), but I think 150 L/h is a reasonalbe requirement .. 1/10 of what D5 is rated. CLCs that to publish their flow rates flow 40 - 96 L/h .. that is 2.6 - 3.0 watts of power motor rating compared to D5's 23 watt motor rating.

So let us look at the highest rated CLC pump compared to D5 sereies pump. ID Cooling is only one giving lift rating.
D5 . . . . . . . 23 watt / 1500 L/h / 3.9m H2O
ID Cooling
Frostflow . . 3.0 watt load / 96 L/h / 1.2m H2O​
That is approx 8x wattage, about 15x L/h & 3x lift.

To me that is a huge difference .. so even if D5 series supplies twice as much as we may ever need, that is still 7.5 times as much as the best CLC pump is capable of.

The are only marginally capable of doing the job, and anyone who has removed one after 2-3 years and placed it with a new one know the new replacement cools much better .. and the only thing that wears out is the pump. The radiator, fans and waterblock are not loosing the performance ability, but pumps wear, and anything but a new LCLC pump is degrading the CLC's cooling performance because of it's lower flow rates.

So I stand by my statement of
'they manage to flow enough' rather than 'they still work'
I believe that 40-60 L/h is 'Just barely working" and as soon as that pump starts being used it starts wearing and flowrate starts dropping. I believe a pump rated 150 L/h is a reasonable lower limit. Maybe even a 96 L/h is enough, I don't know, but D5 series (pretty much an industry standard) is rated 1500 L/h I don't think 150 L/h minimum is unrealistic.
 
Last edited:
For me it's is nothing to do with 'snobbery' and everything to do with function. I agree many probably think they need way more L/h than they need, Most brand new CLCs are about equal to top air coolers. After a couple years use their performance is below that of top air coolers. A good custom loop is significantly better than top air cooler both when it is new and when it is 10 years old.

As an engineer you should understand the difference between 'just enough' to get by and more than just enough so as heat increases more coolant can flow through waterblock keeping it cooler .. or as pump wears the needed flow can be maintained. Head is important because it translates into pump's ability to overcome friction / resistance within the loop .. and as an engineer you should know liquids flowing through hose, waterblock and radiator cores are fighting friction. ;)

I agree, few of us need the kind of flow rat a D5 series pump is capable of (up to 1500 L/h), but I think 150 L/h is a reasonalbe requirement .. 1/10 of what D5 is rated. CLCs that to publish their flow rates flow 40 - 96 L/h .. that is 2.6 - 3.0 watts of power motor rating compared to D5's 23 watt motor rating.

So let us look at the highest rated CLC pump compared to D5 sereies pump. ID Cooling is only one giving lift rating.
D5 . . . . . . . 23 watt / 1500 L/h / 3.9m H2O
ID Cooling
Frostflow . . 3.0 watt load / 96 L/h / 1.2m H2O​
That is approx 8x wattage, about 15x L/h & 3x lift.

To me that is a huge difference .. so even if D5 series supplies twice as much as we may ever need, that is still 7.5 times as much as the best CLC pump is capable of.

The are only marginally capable of doing the job, and anyone who has removed one after 2-3 years and placed it with a new one know the new replacement cools much better .. and the only thing that wears out is the pump. The radiator, fans and waterblock are not loosing the performance ability, but pumps wear, and anything but a new LCLC pump is degrading the CLC's cooling performance because of it's lower flow rates.

So I stand by my statement of
'they manage to flow enough' rather than 'they still work'
I believe that 40-60 L/h is 'Just barely working" and as soon as that pump starts being used it starts wearing and flowrate starts dropping. I believe a pump rated 150 L/h is a reasonable lower limit. Maybe even a 96 L/h is enough, I don't know, but D5 series (pretty much an industry standard) is rated 1500 L/h I don't think 150 L/h minimum is unrealistic.
Yep, I know all those things. The other thing I know, as an engineer, is cost. There's a reason closed loop coolers can be had for around a hundred bucks, and a CPU-only custom loop will cost at least twice that much if you buy only the necessary parts of decent quality. CLCs occupy the engineering sweet spot of maximum efficacy: the greatest performance for the lowest dollar amount. As your numbers show, our custom cooled rigs are seriously over-engineered. We happily pay for that over-engineering and it brings benefits, but it costs way more than it needs to for what it does. The performance per dollar return on investment becomes marginal pretty quickly when you go custom loop, but we do it because it's fun.

Look,

IMG_20170708_113232702.jpg


This is my rig. I'm definitely not here to make the case that CLCs are equal to to custom loop watercooling. I've used both and will likely never put a closed loop cooler in my system again, unless really strapped for cash. I feel the same way about them as I do consoles: they aren't as good as the more expensive option, but they work, they're economical and they have a market. I'm not in that market, but that doesn't make them universally junk.

Sorry for the rant.
 
Last edited:
Yep, I know all those things. The other thing I know, as an engineer, is cost. There's a reason closed loop coolers can be had for around a hundred bucks, and a CPU-only custom loop will cost at least twice that much if you buy only the necessary parts of decent quality. CLCs occupy the engineering sweet spot of maximum efficacy: the greatest performance for the lowest dollar amount. As your numbers show, our custom cooled rigs are seriously over-engineered. We happily pay for that over-engineering and it brings benefits, but it costs way more than it needs to for what it does. The performance per dollar return on investment becomes marginal pretty quickly when you go custom loop, but we do it because it's fun.

Look,

View attachment 29876

This is my rig. I'm definitely not here to make the case that CLCs are equal to to custom loop watercooling. I've used both and will likely never put a closed loop cooler in my system again, unless really strapped for cash. I feel the same way about them as I do consoles: they aren't as good as the more expensive option, but they work, they're economical and they have a market. I'm not in that market, but that doesn't make them universally junk.

Sorry for the rant.
What are you basing our claim
"CLCs occupy the engineering sweet spot of maximum efficacy: the greatest performance for the lowest dollar amount."
Because my on study indicates it's more to do with marketing and how much profit can be made then the 'maximum efficacy' and 'lowest dollar amount'. CLCs also cost more then top air coolers do, have similar performance to noise levels, and if high performance fans (making as much noise as CLC fans) are installed on most top tier air coolers the air coolers cool just as well or better than the CLCs. I agree that custom loops are expensive propositions, but in my opinion CLC performance is simple to close to failure .. and this is backed up by all the RMA and pump problems users have had in the last 6 or so years. Warranty is not an idicator of how long a product will last as much as it's a time limit the companies think is good advertising and numbers they feel comfortable replacing. Many do not bother with warranty anyway. LIke I said before, we are seeing many CLCs now being replaced with top air coolers and these users are all raving about how their 2-3 year old CLC was much louder and did not cool as well as new air cooler is.

Nice rig! I used custom loops for years but went back to air about 5 years ago .. NH-D14 / Silver Arrow time. ;)

No worrys about the 'rant'. Makes the conversation interesting.
 
Last edited:
Back
Top