Test Setup

Cooling Test System
Case Corsair Graphite Series 600T
Processor Intel Core i7-2600K (4x3.4GHz, 3.8GHz Turbo, 32nm)
Motherboard ASUS P8P67 Pro (BIOS version 1502)
Memory G.SKILL Ripjaws X Series 8GB (2x4GB) DDR3-1866 Kit
Graphics Card MSI GTX 580 Lightning
Solid State Drive OCZ Agility 2 120GB
Hard Drive Seagate Barracuda 7200.11 1.5TB
Power Supply Corsair HX850 Power Supply
Operating System Microsoft Windows 7 Professional 64-bit
Test Coolers Stock Intel HSF
Corsair H60/H80/H100
Thermalright Silver Arrow
Thermal Compound Arctic Cooling MX-2

Testing Procedures

For temperature testing we load our Core i7-2600K's four cores and eight threads using Prime95 version 26.6. For the sake of consistency, we use the Small FFT stress test. Most likely in daily use, your CPU will rarely reach the temperatures Prime95 and other applications used for stress testing will produce. This should present a worst-case scenario for CPU temperatures and allow us to really see how these CPU coolers handle heavy loads. Due to fluctuations in the ambient temperature during testing, albeit minor, we use delta temperatures to compare results.

Real Temp version 3.67 is used to monitor and log temperatures with samples taken once every second. We run each test for 15 minutes, take the average temperature of all four cores, and use a three minute rolling average to calculate the results. The final temperature is the average of the last three minutes from the test. We then take the final calculated temperature and subtract the room’s average ambient temperature to get our delta temperature. This method has the least temperature fluctuations and is the most consistent.

In addition to the Corsair Hydro Series coolers, we’re testing with Intel’s stock HSF as a baseline, and representing high-end air-cooling is the frankly massive Thermalright Silver Arrow. The Silver Arrow comes with two 140mm fans, with one fan sandwiched between the two large radiator towers. It also weighs in at a hefty 825g without the fans, or around 1.2kg with the fans and clips. Thankfully, it comes with a good mounting solution so as to avoid putting too much strain on your motherboard, but there’s no denying the fact that this is a heavy cooler.

Each cooler is mounted and retested three times to verify good contact was made. We use Arctic Cooling MX-2 instead of the thermal interface material (TIM) that comes preinstalled on the test CPU coolers. There is no curing time allowed between mounts or changing CPU coolers—the MX-2 TIM we use claims not to need curing anyway. These methods help ensure consistency across all tested coolers as well as provide comparable results.

We test each cooler at the Core i7-2600K's stock 3.5GHz frequency, which is the speed it runs at if all four cores are maxed out at 100% and Turbo Boost is enabled. For testing purposes, we disable Turbo Boost and manually set the clock speed to 3.5GHz, with Hyper-Threading enabled; the stock speed runs at 1.16V. We also perform the same set of tests with the CPU overclocked to 4.8GHz using 1.4V, again with Turbo off and Hyper-Threading on.

For noise measurements, we use a Check Mate CM-140 SPL meter. All noise tests are conducted between 1 and 3 AM to ensure the lowest possible ambient noise. In our test environment, we measured ~29 dBA with the test system turned off. Before and after each noise test, the same ambient ~29 dBA measurement was verified. We measured 1 foot away from the test chassis with all doors installed. Our goal is that this method for measuring noise will best mimic a typical usage scenario, though obviously the choice of case, power supply, and graphics card also plays a role.

Cooler Installation Temperature Results


View All Comments

  • Mjello - Monday, November 07, 2011 - link

    If you don't mind temps howering in the 70C-75C range the h100 can do it passive mounted externally in a horisontal position in a 25C room.

    Just did an hour of LoL. It wont do prime95 for more than 10 min. however. So heavy loads is a nogo passive. And a silent fan doesn't make any noice so i'll switch mine back on. Was a fun experiment though.
  • double0seven - Tuesday, November 08, 2011 - link

    "We could hook up our video cards and CPU to the loop."

    This part doesn't make logical sense. If you "hooked up" multiple heat generators to a coolant loop, only the first one would get cooled.
  • ggathagan - Monday, November 07, 2011 - link

    Page 3, last sentence of the 1st paragraph: I didn't have any filament problems....
    I'm guessing fitment is the word you intended.

    Hey, someone has to pick at this stuff...
  • fixxxer0 - Tuesday, November 08, 2011 - link

    what was the ambient temp of the room on average? just curious to see what your actual temps were... Reply
  • compudaze - Tuesday, November 08, 2011 - link

    Ambient temps were 22.2C - 24.8C with an average of 23.8C. Reply
  • medi01 - Tuesday, November 08, 2011 - link

    It's been a while, that most power hungry part of the system is a graphic card.
    What a pity they don't make water cooling solution for them.

    I don't quite get why on earth I would install this thing, if I still have to care what to do with graphic card heat.
  • kg4icg - Tuesday, November 08, 2011 - link

    Cool IT does make a sealed solution for video cards.

  • Sabresiberian - Tuesday, November 08, 2011 - link

    It's good to see these coolers get to the point where they are really good, even if the cost is a bit high still. I think pretty soon, liquid coolers will be the thing to use in all but inexpensive systems.

    I'm not sure the H100's price is too high; it depends on what a serious overclocker wants, and whether it allows more room than something like the Silver Arrow. Adding another 2 fans will make even more difference (and add to the cost), which will give more of a performance edge to it. Theoretically, anyway, air flow through the radiator may not be the limiting factor in the system.

    I use the CoolIT Eco on one of my rigs (similar to the H60). It's a decent cooler, but the main reason I used it was because I needed a lower profile solution with my Antec Skeleton "case", and these type coolers really work great for that. Of course, you are just moving the space issue to a different location, since there's the radiator to deal with, but that's not an issue with the Skeleton since I just made a bracket to hang it off a side.

  • Googer - Wednesday, November 09, 2011 - link

    Shame corsair didn't show off the "Corsair Link" kit and demonstrate its abilities. I guess they don't want to sell too many of those $100-129 (USD) kits. Sounds like the Bob's Rug Syndrome (GTE)
  • Drekkyk - Wednesday, November 09, 2011 - link

    I tried the H60, but at 4.5 Ghz the temp was 8 degrees C hotter than my Cooler Master Hyper 212+. under load and at idle. What I really noticed was how much slower the temp reduced when going from full load to none. I really wanted to like it, I even tried remounting, ensuring thermal compound was applied correctly, etc. I used the same compound for both coolers. My case is a Haf-X with plenty of airflow. Reply

Log in

Don't have an account? Sign up now