Anyone that has listened to this show knows that I am a huge fan of the self contained water cooling systems we can now get. They offer cooling performance at or near the level of the high end air coolers but do it with less noise and a lot less space taken up inside the PC. This last point is in fact one of the major selling points for me as it means that it is easier to work inside your PC with the cooler installed and leaves all of your RAM slots free for maximum memory potential.
The beginning of this style of cooler was pretty simple, a radiator, and fan with the cooling block, if you needed more cooling potential you just added a second fan. The second round saw an increase in the radiator thickness to allow for more cooling surface are and the second fan became standard. This second generation never made any sense to me. I mean I understand the reasoning but it never made any sense because there seemed to be a better option available, well enter the H100 and that better option.
Corsair switch teams recently and moved from Asetek to Coolit for it’s partner self contained liquid cooler design market. Coolit made a move as well it seems, withdrawing from the consumer market to focus on it’s OEM work as well as working with partners like Corsair. This meant that some of Coolit’s designs became available for Corsair to make use of and the 240mm radiator model was one of those designs.
The idea is actually pretty simple, if 120mm of radiator can do a good job at cooling the 240mm should do even better. Now unlike the second level of coolers this is not a thicker radiator, it is a longer radiator. The setup will accommodate 2x 120 mm fans or 4 fans if you want to use a push/pull configuration. The 240mm radiator was already pretty standard fair among the custom liquid cooling crowd and a good number of cases came with support for this size radiator in place. That is why I wonder why it took so long for this design to come out and why more are not doing it.
The move to Coolit was a great one for Corsair in my opinion because while I think Asetek produces great products their mounting system is something I am not a huge “fan” of. However if you go back to my reviews of the CooliT Eco you will see I love the simplicity of the Coolit mounting system and this carries over to the H100. Once you have the back plate in place the mounting is done using 4 thumb screws. No messing around with rings and rotating things and lifting while position and such. Just put it in and screw it down, tool less if you like.
The radiator is attached to the cooling block via two hoses to move the liquid. This is the one area I wish Corsair and Coolit would modify a bit. They use an outer cover over a thick rubber tube. The outer cover is meant to provide additional protection and it does but it also makes the tubes less flexible that other systems. This is not to say the tubes will not flex, they will do the job but you might have to apply a bit more muscle to get them to run the way you want them too.
The two included fans attach to the cooling block rather than the motherboard for their power and speed control. Both are 4 pin PWM fans so we have a nice range of speed options. The cooler itself is attached to the motherboard CPU fan header as well as to a molex from the PSU for additional power. The connections on the cooler allow for up to 4 fans and this is just to much power draw with the pump itself for that single header.
The cooler itself has an LED display on it to show which of the three possible fan modes you have active. Mode 1 is to set the fans to lower speeds and thus quiet operations. Balanced mode with a fair low bottom speed and a good range to the top speed and then Mode 3 which ranks everything up. I did some digging, surprisingly this was hard to find with any kind of official material and found this chart concerning the speed ramping of the H80, this also applies to the H100.
The chart is supposed to have come from a Corsair blog entry and I have not had much luck finding it but then again after a bit I quit digging. Based on my testing the chart seems about right. You can see that all three modes ramp up the speed as the coolant temp within the system gets warmer. Each level up raises the base speed and max speed of the fans and thus increases the noise.
So how does it actually cool? Well we put it to the test using the Thermaltake Level 10 GT case and an i5 2500k processor. I wanted to get some load on the coolers so we overclocked the processor to 4.1 GHz and started testing. For comparison purposes I used a Thermaltake Frio as well as an Antec H60 cooler.
Using Prime 95 I fired up a 30 minute load on the processor and took temperature readings from each of the coolers. I tested the Frio at it’s lowest, roughly middle and highest settings, the H60 at stock settings and the H100 at Quiet, Balanced and Performance. Even on the Quiet setting the H100 started strutting it’s stuff out of the gate delivering a nice 3C difference over the H60 at stock and the Frio on low. Cranking the H100 to Balanced settings moved the H100 a bit further from the pack going to 5C cooler than the Frio at mid fan speeds and 6C cooler than the H60 at stock settings. Finally on high the lead was solidified with the H100 dropping a full 7C below the Frio and 8C below the H60. That’s some serious cooling.
Now I can hear some question rising in the ether already. Why did I not mention idle temps? The reason is simple ANY cooler can do a decent job at idle and these three are great coolers, the idle temps where so close as to not be an issue worth measuring. Why did I not give actual temperature readings and only the differences? The reason is the actual temperature readings are meaningless, I guess I should explain that. The actually temperature readings anyone will achieve are based off a large number of factors. These include the room temperature, air flow restriction of case location, the case and fans used in the case and the list goes on. This means that just because I hit a certain temperature point, you will not automatically hit the same point. As such the differences in temperatures for each cooler is more meaningful since it gives a comparison that should be close to the same for everyone.
At $102 on Newegg the H100 is very much priced at a luxury level component but you might be surprised to find I can see a value option in it. You see a lot of cases today and even some of the more value priced cases have the ability to use a 240mm radiator in them. With this in mind a lot of the people that will be building tweaker systems will be putting in the dual 120mm fans and then adding a good cooler to the mix. The H100 allows the potential to kill two birds as it where with one stone and be greater than the sum of the parts. When you figure the cost of the 2 fans and the nice cooler and compare it to the cost of the H100, it suddenly does not seem that bad.
I do have to give a warning here, not all cases that will fit dual 120mm fans will handle this cooler. There is of course the spacing that needs to be set for the radiator to mount but that can be worked with. Of bigger concern is the space for the radiator and fans to sit. If you case would be a tight fit with just the dual fans in place then this cooler is not going to work for you. You need to check the spacing at the top of the case before pulling the trigger on this cooler.
If however you have a case that will accommodate this and you are in the market for a new cooler the H100 should be at the top of your list. With just the stock fans this brute has cooling power to spare. Add to this the ease of installation of the Coolit design and the ability to profile your fan ramping to meet your cooling needs, this is a great CPU cooler.
Corsair H100 Review as it aired 28 August 2011