Quote:
|
Originally Posted by murray13
GREAT article!
Only drawback is that it makes an assumption that you are only going to water cool the processor. Personally I would like to see the numbers for a full system being cooled. Example: CPU, NB, GPU, and maybe even a HD in the loop as well. The only other thing I would have liked to see is the numbers for the 30RZ at 60Hz. It would put things into a little better perspective for those of us here in North America. It still only adds about 50W of heat but has a little better numbers.
Just my two cents.
Thanks for bringing it over here Stew, not all of us get to OCAU all the time.
Dave
|
All your suggestions are already on the cards for the expanded article, with the exception of NB and HD blocks, as it is my very strong opinion that cooling either is next to useless. With the data I'll be presenting, it should be pretty evident that such things don't really push the final results around all that much, as in there is a pretty strong case across a very broad range of scenarios that suggests that there is a fairly level plateau with respect to what makes a "good pump". I mean take a look at the Eheim 1048 vs the Iwaki MD-20RZ, there's hardly much difference (1.2C), and a couple of additional blocks don't suddenly chop flow rates in half, and even if they did, it still doesn't really affect the final rankings in a dramatic fashion.
What has the MOST impact on altering the final rankings of which pump is best for any particular person's scenario is in fact the radiator and the amount of cooling power it offers. This is the single largest impact item for determining which pump is best for someone's scenario, and not so much which blocks they have in the loop.
In any event, this will all come out in the full article that I'm working on. I'm still busy gathering more data and generating more graphs....