Newbs question on CPU block

[unclecheong]

I did some search on the web about the performance of CPU waterblock. I find a water block rankings (http://www.watercoolplanet.de/index.php?open=4&show=1&pagenum=1) by Watercoolplanet which listed 47 commercial available CPU waterblock.

This is writen in german hence I'm not totally understand :( . I doubt the result because the CASCADE is one of the best waterblock why it is outperformed by the germany one so much.

Is the test reliable or not?

Many thanks. ;)

[Bevman]

Interesting. Notice that almost all the top 10 spots are using some sort of knock off of the Cascade's design.

1 Cyclone EvolutionS better than RBX?
2 RBX better than Whitewater?

Don't know if i believe that.

[johan851]

Those tests were done with European setups - very, very low flowrates in comparison with the systems we generally use. Since blocks like the RBX, WW, and Cascade all require high flowrates to perform properly, it's no wonder that the other blocks built for low flowrates win. If I put one of those 'winning' blocks in my system though, it would choke. Different strokes for different folks.

THIS is what I would consider absolutely reliable, and more up your alley:
http://www.overclockers.com/articles373/wbsum.asp

[Cathar]

Many of the "English-speaking" derived blocks are quite thin-based, with base-plate thicknesses well under 2mm thick. The Cascade is 0.7mm thick, the White Water is the same, and the RBX averages about 0.95mm thick with its pits. The Silverprop blocks are of similar bp thickness as well.

This presents an issue when it comes to testing. These ultra-thin base-plated blocks absolutely demand that they be matched with pumps of adequate power, with adequate being at least an Eheim 1048 in an otherwise unrestrictive setup as a minimum, or an Eheim 1250 as a recommended sort of pump. Of course the blocks can be run down to well below their designed operational flow-points, but that's not really being fair to the intent of the design.

The WCP testbed uses an Eheim 1046, and couples that with a very high restriction radiator. Given the setup, the flow rates of the testbed would be around the 2LPM (0.5GPM) mark before the added restriction of the tested waterblock.

Many of the results at WCP also do not gel with multiple other tests of these same blocks at low-flow. The ordering is all wrong, and some of the differences between blocks are vast. This is also due to WCP only conducting a single test of a water-block, with the subsequent up to 3C variations that can occur with just a random mount of any block.

Basically we have an test-bed with an uncontrolled procedure running blocks at flow-rates significantly below the levels intended for many of the designs being tested.

Go take a good look at Procooling.com tests by Phaestus, and the OC.com tests by JoeC. Phaestus offers and avantage over the OC.com tests by providing a flow vs performance curve, and offers results of multiple mounts, and tests at flow rates right down to the levels that the WCP test-bed operates at. The results at OC.com and Procooling.com do not agree with WCP in even the slightest fashion.

Comes down to the individual on which testbed and procedure that they want to believe. Both Phaestus and OC.com attempt to control the mounting variables. WCP do not. Both Phaestus, and to a lesser extent OC.com, attempt to offer results at "typical" flow-rates for the blocks tested. WCP does not.

You choose what to believe.

[Bevman]

Thank you Cathar