Snazzy Water Block Roundup

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
55,535
This site is in German, but we have linked it through Google's translate here for your English reading pleasure. ComputerBase covers five high end blocks built for custom cooling loops. Besides just running temperature benchmarks, they tear the blocks down so we get to see the goodies inside.

Der Kühler erlaubt sehr hohe Durchflusswerte und zeigt eine ordentliche Kühlleistung, kann aber nicht ganz mit dem Testsieger mithalten. Für Kunden, die Wert auf die Beleuchtung legen, ist er dennoch eine Überlegung wert. Wer sich an der etwas geringeren Leistung stört, kann diese durch mehr Radiatorfläche und damit kühleres Wasser problemlos kompensieren.
 
Here's what I use for my custom water loop. It's a water block, not an overpriced iPhone.
I don't disagree, but my XSPC block wasn't ridiculously priced either--only $50 on performance-pcs or amazon, $20 if you get it used from amazon. Considering the increased fin-density (they appear to be rather thick fins from the ebay photo, hard to tell through the acrylic), I think it's worth the minor bump in price. Edit: Looks like it's on Amazon, too. As I suspected, the fins are pretty large and have pretty big gaps between them. It'd work okay compared to a raystorm at a low flow rate, but at higher flowrates the raystorm would dominate.

If you're strapped for cash, I wouldn't blame you for going that route, though.
 
Last edited:
As an eBay Associate, HardForum may earn from qualifying purchases.
As an Amazon Associate, HardForum may earn from qualifying purchases.
I've been wanting to switch to watercooling... Any recommendations for CPU and dual GPU?
 
Interesting. The EK Supremacy seems conspicuously absent, especially since it seems to take the performance crown in every other review I've read.
 
I've been wanting to switch to watercooling... Any recommendations for CPU and dual GPU?
Get what you can afford, make sure you have space for the amount of radiators you'll need to cool all three. If you'll be overclocking and you can afford it, full-cover blocks would be best. The only GPU block I have first-hand experience with is Koolance's GPU-220, which seems to do pretty well, but the mounting mechanism is a bit interesting. For CPU blocks, there's mostly minor differences, and it'll depend on the speed of the water through your loop, but EK's Supremacy and Watercool.de's Heatkiller are top end, whereas XSPC Raystorm is the budget champ. Else, you can get the cheap chinese ones if you just can't afford XSPC, but you probably won't get much better performance than a good air cooler with one of those.
 
Precision milling equipment isn't cheap. Skilled workers to use precision equipment isn't cheap. R&D isn't cheap. Shut up, you got exactly what you paid for, a cheap hunk of copper.

Milling a waterblock isn't that precise. A water block is a hunk of copper that water flows over, the difference between the best of them and a cheap one isn't going to be huge. I'd say for most water-cooled pc's you wouldn't need it. I'm running an old dtek fuzion v2 that I got used off of [H] and it does the job. Even an expensive waterblock is still just a 'cheap hunk of copper'.

R&D isn't cheap and that is one reason why premium waterblocks can cost a lot, but its also because if you spend some time on aesthetics and marketing the market will bear the price because what is a $60-$100 waterblock in a $2k+ build?

I don't totally disagree with you, I'm sure the cheap ebay one is worse, maybe even by quite a bit seeing as how cheap no name Chinese things are more about margin than doing their job, but there is no reason to be pissy about it. If he bought a 'cheap hunk of copper' that does its job and doesn't cost extra for looks/name then why not recommend it?
 
The D-tek Fuzion V2 has been shown to be about 5 C worse than the XSPC Rasa, and today's top blocks are at least 4-5 C better than the Rasa last I recall. In any case, your "cheap" Fuzion was way better machined than the piece of junk dukenukem got. His piece of junk might marginally do the job, but even an old Ford Model T can get you from point A to point B. It won't do it as well as newer cars can. His comparison of CPU waterblocks to iPhones was poor, to put it mildly.
 
Milling a waterblock isn't that precise. A water block is a hunk of copper that water flows over, the difference between the best of them and a cheap one isn't going to be huge. I'd say for most water-cooled pc's you wouldn't need it. I'm running an old dtek fuzion v2 that I got used off of [H] and it does the job. Even an expensive waterblock is still just a 'cheap hunk of copper'.

Wrong. They aren't just hunks of metal. The best performing blocks have various specialty patterns or fins mulled into the surface to maximize the heat transfer between the blocks and the fluid. The Supremacy EVO even has a set of exchangeable nozzles and jets to optimize the block for different CPU designs.
 
Get what you can afford, make sure you have space for the amount of radiators you'll need to cool all three.

I need multiple radiators? For some reason I was picturing a big single radiator being enough..
 
I need multiple radiators? For some reason I was picturing a big single radiator being enough..
Not necessarily, should have said "radiator(s)". You may need multiple if your case is small--for example, if you only have room for a 2x120mm radiator up front, you may want another 120mm radiator in the back or consider getting a larger case. A 2x120mm radiator might be enough for two gpus + cpu if you weren't overclocking, but if you are then it'd probably be a good idea to get a 3x120, 2x120 + 1x120, or 2x140 radiator (minimum).
 
Wrong. They aren't just hunks of metal. The best performing blocks have various specialty patterns or fins mulled into the surface to maximize the heat transfer between the blocks and the fluid. The Supremacy EVO even has a set of exchangeable nozzles and jets to optimize the block for different CPU designs.

That doesn't make it not a 'cheap hunk of copper' I mean yeah its a hunk of copper cut in special ways but it would be wrong to say I'm flat out wrong. To add to that even the cheap waterblocks have fins cut into them unless you get those sub $10 literal hollow hunks of copper blocks. Just because they haven't done any R&D on restriction/flow rate/etc doesn't mean its not close performance wise, its still water conducting heat from copper, so long as there is sufficient surface area and flow rate it'll be pretty good.

Silly argument anyways, I conceded that more expensive waterblocks do have R&D done to them in the same post you quoted.
 
Last edited:
15 dollar block around thousands of dollars of hardware is a bit risky in imho. Use only ek stuff here, or major brands for things like this, now not saying the chinese one is junk, how do ya say unknown lets call it that...
Anyone had it for years? Tested it or reviews?
 
That doesn't make it not a 'cheap hunk of copper' I mean yeah its a hunk of copper cut in special ways but it would be wrong to say I'm flat out wrong. To add to that even the cheap waterblocks have fins cut into them unless you get those sub $10 literal hollow hunks of copper blocks. Just because they haven't done any R&D on restriction/flow rate/etc doesn't mean its not close performance wise, its still water conducting heat from copper, so long as there is sufficient surface area and flow rate it'll be pretty good.

Silly argument anyways, I conceded that more expensive waterblocks do have R&D done to them in the same post you quoted.

The good blocks have fins/pins of minimal thickness and gaps to maximize surface area. Those cheap blocks with 3 fins cut into them do not maximize surface area.

DT tried something unique a few years ago. It was literally putting the water onto the IHS of the CPU with no copper block in-between. Conclusion? It performed worse than normal blocks, because those normal blocks had all the microfins/pins to facilitate heat transfer, which the IHS does not. The R&D into that fin/pin structure for heat transfer alone is not insignificant, let alone restriction and the rest.
 
DT tried something unique a few years ago. It was literally putting the water onto the IHS of the CPU with no copper block in-between. Conclusion? It performed worse than normal blocks, because those normal blocks had all the microfins/pins to facilitate heat transfer, which the IHS does not.


That shouldn't be surprising. It is the water equivalent of running a CPU on air without a heatsink.
 
The good blocks have fins/pins of minimal thickness and gaps to maximize surface area. Those cheap blocks with 3 fins cut into them do not maximize surface area.

DT tried something unique a few years ago. It was literally putting the water onto the IHS of the CPU with no copper block in-between. Conclusion? It performed worse than normal blocks, because those normal blocks had all the microfins/pins to facilitate heat transfer, which the IHS does not. The R&D into that fin/pin structure for heat transfer alone is not insignificant, let alone restriction and the rest.

no fins != cheap water block

It really isn't surprising that a flat surface didn't have as much surface area to conduct heat through.
 
Back
Top