Best CPU Cooler for i7 4770k Haswell?

How about the Phanteks PH-TC12DX? It's not as big, it will work at your OC levels, and it is quiet. If you get a color that works with your theme, it can look great too because they do come in a variety of colors.

Cooler you said is ~$20 more than a 212evo, and it only preforms a max if ~5C hotter, often the same, or an even smaller difference. As such, I don't think either of them can hit 4.5, I think 4.2 is the max if you have a good chip. My friends could barely hit 4 on a 212evo(his chip was shit), so if you want to hit the higher end of your range, you'll need a closed look like a h100i, presuming you refuse to use a bulkier air cooler.

@Dylan Patel: It is not very big compared to the Noctua NH-D14 or the Phanteks PH-TC14PE. So if motherboards can fit those in, they can surely support the smaller one.

The CLC coolers are not actually maintenance-free. Air coolers are maintenance free. The reason why CLC are not maintenance-free is because they can degrade over time. And since they all have non-transparent black tubing, you can have a hard time noticing. Best way is to keep a monitor of your temps.

You could argue that air also isn't maintenance free, you have to dust it, and replace thermal paste every few years. I've never really heard of h100is degrading, there have been a few horror stories of them leaking, buts its still pretty rare. I went air cause it was cheaper, and I was a bit paranoid.

@googoo1876: if you dust filter correctly, it will be fine. Also, the fans are the ones to accumulate dust. The heatsink itself rarely accumulate dust. And if so, it isn't much so just blow at it.

I don't quite understand why you have to refill the loop ? Do you plan on opening it up and modding it ? If you want performance and low noise the H220 is a good buy.

ok i am pretty blind and saw this post. I am going to give you the full answer to this, so forgive me if i ramble a bit

I am a chiller/refrigeration mechanic. Its my job to know about saturation points, superheated vapours, subcooled liquids and all sorts of nonsense that noone but us really care about. My highest level of education in this area is refrigeration tech colledge, yr 12 chemistry & physics, & a 6 month stint at uni in electrical engineering that i truely hated. I say this so if you are a chem major or something, please dont crush my vocab or reasoning. Also I'm Australian, so i work in metric for temperatures - cant work in fareiheit sorry

Lets start by saying water is a special compound. It has a very wide liquid vapour saturation point. A 100 degree (C) saturation point. What does this mean? Lets start with the basics:

There are generally 3 states to MOST elements and compounds:- solid, liquid and vapour. Sure some only have 2, skipping liquid and going directly from solid to vapour, and then there is a fourth state that some have called plasma, but that is very uncommon. Lets just say that all "stuff" had 3 stages. At low temps/pressures element x is a solid. As you add energy (heat), the temperature of element x rises until it hits a point where it changes state - this is called the saturation point. Now this point is normally very small and is generally just given a degree value. Since were are talking about water, lets use that example. Waters liquid solid saturation point at normal atmospheric pressure (ie. at sea level) is 0 degrees C. This is generally what we call the freezing point of water (32 F or whatever you yanks call it ) So we add heat to our minus 10 degree ice cubes and then bam we hit the magic number and all our ice is water right? No. At zero, H2O can be both water and ice. if the ice was minus 1, it would HAVE to be ice, and at 1 degree it would HAVE to be water, but at it's saturation point it can be both, because at that temperature it is in the process of changing states (its the difference between latent and sensible heat, which i wont get into.

OK, got it, so what about this 100 degree saturation point if water solidifies at a fixed temp? Well, waters boiling point is 100 degrees celcius so isnt that it's liquid vapour saturation point? Nope. Waters LV satpoint is from just above freezing to just bellow boiling. Water can exist in those states for a very wide range.

OK WTF I wanted to know about watercoolers, not all this other crap. Well it basically means this:- Since water can be a fluid and a vapour at the entire temp range of your spiffy little water cooler, some of the water will evaporate. On an open circuit cooler, it will evaporate from the reservoir, and you just fill it up - easy. In a "closed circuit" cooler it cant do that. You know all those joints in your piping - they will be 100% water tight. Great. BUT! Because the circuit is closed, the loop is presurized because water expands as it takes the heat of your CPU. OK so its still a closed loop so it should stay in the the tubes right? If the whole thing was a properly sealed system with copper piping and braised joints, then yes, the vapour would stay and eventually recondense into a fluid when it cools. But its not a properly sealed system - it has plastic tubes and clamps and a bunch of other easy to assemble pieces, so that while they may stop fluid water from leaking out and frying your system, it will not 100% stop a pressurized gas/vapour from leaking out.

Now this (little?) rant was only concerning the water evaporating and escaping - it doesn't even touch on what happens to the plastic tubing and other components gradually breaking down from heat and becoming a sludge in the system (if you opened a 3yr old cooler and saw what the "water" looked like you would definitely not drink it), or that the water in the cooler is not 100% pure water so it will be filled with all types of minerals that will react with the oxygen in the water and cause other sorts of issues (like electrolosis).

What this all means is
1. Your closed loop water cooler is not really a closed loop after all.
2. you want to clear that crap out as often as you would clean the dust of your heatsinks and fans.

/drunkenrant