Our Water Cooling 101 series is brought to you by our great friends at Aquatuning! Check them out for all of your water cooling needs!
There are many reasons we get into playing around with computers. For some of us it’s gaming, for others it’s pure curiosity, or maybe even it’s because of work. How we come to be enthusiasts doesn’t matter much as we all share in this awesome obsession with tech and computer hardware. Maybe you’re new to our little world or perhaps you’ve been around for years. But for those paying attention there is really just one thing that is most important above all with computing: speed. The faster our machines run the more lines of code can be executed and the more we can accomplish; in turn the better our games and apps perform.
There are a couple of ways to make your rig faster. The first and most obvious is to buy faster parts. There is always a newer, better CPU or GPU coming out and then of course there are better motherboards and faster clocked ram. Whether you choose Intel or AMD there are both enthusiast parts and the bigger, badder, faster enterprise parts such as those found on Intel’s X99 platform.
These parts typically have more cores than the enthusiast parts and are often faster chips. For some of us, probably most of us, just buying new parts all the time isn’t really an option. Or what about just needing a little bump to run that newest AAA title at the max graphics settings? And of course there are those who have the fastest meanest parts on the market and still want to run them faster. This is where overclocking comes in.
By now I have to assume that you have at least heard of overclocking. Thanks in part to smartphones and manufacturers trying to hype up their products the term has become pretty mainstream. Just so we’re on the same page let’s go ahead and spell out what overclocking is exactly. When a processor is released by the manufacturer the speed that it operates at is referred to as the clock speed. This is in reference the speed at which the processor completes one processing cycle. Back in the day it was discovered that by hacking the manufacturer settings one could force a processor to operate at a cycle rate faster than that which it was originally designed. This is not to say that it is beyond the ability of the chip, rather that it is faster than the safe zone that say Intel or AMD was willing to set the chip speed at. While there is risk of damage to the overclocked part, this is generally considered worth it for the performance gains.
Over time this hacking evolved into a much more sophisticated process of alterations to the CPU’s speed and power settings within BIOS’s and now UEFI. In fact the process is now fully supported by almost all modern motherboards and while still slightly risky with just a small amount of research a relatively safe overclock can be applied to a CPU by even a novice user.
I did of course mention risk. Let me expound on that for a minute. When you overclock a processor you are asking it to complete more cycles in a given amount of time. That is, making it do more work in the same amount of time than it was previously doing. Physics requires that in order for this to happen more energy is going to be required in the first place. In the case of all computer equipment that energy is of course electricity. Say we want to take a given processor from 3.0ghz up to 3.5ghz. This is a speed gain of ~15%. In order to do that the processor’s core, the group of parts we are speeding up, is going to need to be given more voltage (energy). While there is no exact science to this (and is a separate subject really) in general it won’t require a 15% voltage gain since there should be a pretty generous amount allotted to the core parts by the manufacturer. However it will be consuming more power regardless of what we do with the core voltage. What this means to you the user is that when your processor consumes more energy there is going to be a by product of the energy use; that product is heat. The extra heat produced if not correctly accounted for will actually cause the processor to hurt itself. Depending on how extreme the overheating is it can cause damage ranging from slow long term degradation to an immediate thermal blow out: a burned up processor.
So what to do? The manufacturer supplied cooler is only designed to cope with the amount of heat produced by the CPU left basically at stock settings. But of course you don’t want to settle with those slow speeds do you? And that is where aftermarket cooling comes in. Over the years there have been a myriad of options made available. These range from air coolers like the one that came with your CPU, to the extremes such as phase change refrigeration units, liquid nitrogen CPU blocks and everything in between.
Air coolers are designed usually with a thick metal base that makes contact with the top of the processor and passes heat through metal pipes up into thin fins of metal that radiate that heat out and away from the CPU. While this is the same principle implemented in the stock cooler, aftermarket coolers just do so with better, more efficient metals and are usually much larger. In almost all products a small fan typically moves air over the fins of the cooler to move the heat away.
While air coolers can do a really good job they are the base level for processor cooling. Next in order of performance we come to a cooler design now referred to as an “all in one” liquid cooler or AiO for short. These units feature a radiator just like the one found in your car connected to a similar block of metal used in the air cooler. Two hoses (think inlet/outlet) are connected to a pump; typically built on top of the metal heat sink that circulates liquid from the heat sink into the radiator where fans move the heat away. This system comes from the manufacturer prefilled with liquid and requires no end user maintenance.
AiOs have recently and quickly become the most popular trend in cooling. They are very cost effective, perform well, and are simple for the end user to use and maintain. It is this simplicity that has made them so popular but also what holds them back. As a product they are evolved from the mother of processor cooling and our main focus: water cooling.
Water cooling has been around for a very long time and has never been more popular than now. This is due to a number of reasons, including how robust water is at dissipating heat. Water has an enormous capacity for quickly taking on, holding, and expelling heat. Coupled with cleverly designed metal heat sinks called water blocks, water pumps and radiators (again just like in your car), water cooling offers you the ability to take your wimpy stock processor to the next level.
A quintessential water cooling system or “loop” consists of a CPU block to absorb and pass on the heat directly from the processor. A radiator which has water channels and very thin fins of metal that radiate the heat from the water out of the loop. A pump to move the water along through the loop. A chamber or container called a reservoir which holds a larger amount of water and allows the loop to be easily filled and supplied with water. And of course some tubing to connect everything together.
Seems simple right? Well that would be the really dumbed down version of what makes up a water cooling loop. There are many other parts that help to make the loop come together. You will of course need some fittings to connect the tubing to the components. These are metal pieces that screw into ports taped to a standard G1/4 size hole. There are many types of fittings that you may or may not need depending on your loop’s design. The standard connector is called a “barb”. This connector goes into the open end of the tubing and features ridges that seal the tubing and prevent it from slipping off the barb. Typically it is recommended to secure a barb with a hose clamp or even a zip tie. In the place of a barb a slightly more secure, but more expensive option, is the compression fitting. This fitting has a nozzle like a barb but also has a metal ring that goes over the tubing and screws onto the base of the fitting. As it screws down it compresses the tubing tightly to the nozzle so it won’t come off or leak. It is superior to the barb in it’s ability to maintain a water-tight seal as well as a typically cleaner look. Additional fittings allow you to achieve angles and reach places you otherwise couldn’t. There is even specific type of fitting called the quick disconnect that allows you to remove a part of your loop without loosing water.
In addition to fittings you will almost always need fans. Many users come to water cooling to try to quiet their rig down from a noisy air cooler. Thanks to the design of many radiators this is possible with some really quiet fans slower speed. Typical water cooling fans are 120mm or 140mm in diameter. Like all aspects of water cooling there are endless options when it comes to fans and their speeds which we will cover more extensively later.
While a typical loop is CPU block, tubing, radiator, reservoir the options do not end there. Depending on the room you have available in your case and your cooling needs/goals, many I would even say most, water coolers have waterblocks for their GPUs as well as multiple radiators or varying sizes and lengths! Beyond that there are waterblocks for your motherboards, ram, chipset, hard drive, and even one for a solid state drive now! Some systems become so extreme waterblocks are even needed for the fan controllers!
Water cooling while extremely effective and often times beautiful to look at can be entirely overwhelming especially if you are new to the hobby obsession. Fortunately we here at ThinkComputers have your back! We will be imparting our years of water cooling knowledge and experience to you in a series we’re calling: Water Cooling 101. Follow along and we’ll help you pick out the best parts for you, show you how to install those parts, and have your own loop up, running, and free of air bubbles in no time.
Welcome to Water Cooling 101!