Comparisons
As mentioned on the individual thermal results pages for the Corsair H50 and Scythe Kabuto, the CPU/case deltas are artificially low, due to some of the airflow through the heatsinks/radiators blowing toward the thermal sensor, therefore please bear this in mind when looking at the CPU/case deltas and case temperatures for these coolers.
| System performance - Idle (Stock) | ||||||
|---|---|---|---|---|---|---|
| Cooler | Arctic Cooling Freezer Xtreme Rev. 2 |
Corsair Hydro Series H50 |
Scythe Kabuto | Thermalright MUX-120 (push-pins) | Zalman
CNPS10X Flex (1-Fan) |
Zalman
CNPS10X Flex (2-Fans) |
| Ambient temperature | 22.7°C | 23°C | 22.8°C | 22.7°C | 23.1°C | 23.1°C |
| Case temperature | 24.7°C | 25.4°C | 24.8°C | 24.8°C | 24.5°C | 24.7°C |
| Ambient/Case delta | 2.00°C | 2.40°C | 2.00°C | 2.10°C | 1.40°C | 1.60°C |
| CPU temperature | 29°C | 28°C | 29°C | 29°C | 28°C | 27°C |
| CPU/Case delta | 4.30°C | 2.60°C | 4.20°C | 4.20°C | 3.50°C | 2.30°C |
| CPU/Ambient delta | 6.30°C | 5.00°C | 6.20°C | 6.30°C | 4.90°C | 3.90°C |
[graph 3105]
[graph 3103]
Despite the issue pointed out at the top of the page, the coolers clearly aren't warm enough to skew the CPU/case delta temperatures in favour of the Kabuto or H50.
The small difference between the first four coolers can be dismissed by the natural variance in our test equipment. However, the CNPS10X Flex in dual-fan configuration is noticeably superior to its rivals and eeks out a 1-2 degree difference at idle.
| System performance - Load (Stock) | ||||||
|---|---|---|---|---|---|---|
| Cooler | Arctic Cooling Freezer Xtreme Rev. 2 | Corsair Hydro Series H50 |
Scythe Kabuto | Thermalright MUX-120 (push-pins) | Zalman
CNPS10X Flex (1-Fan) |
Zalman
CNPS10X Flex (2-Fans) |
| Ambient temperature | 23.3°C | 23.9°C | 23°C | 23°C | 23.7°C | 23.8°C |
| Case temperature | 26.5°C | 28.8°C | 30.7°C | 26.4°C | 27°C | 27.3°C |
| Ambient/Case delta | 3.20°C | 4.90°C | 7.70°C | 3.40°C | 3.30°C | 3.50°C |
| CPU temperature | 59°C | 54°C | 60°C | 55°C | 58°C | 56°C |
| CPU/Case delta | 32.50°C | 25.20°C | 29.30°C | 28.60°C | 31.00°C | 28.70°C |
| CPU/Ambient delta | 35.70°C | 30.10°C | 37°C | 32°C | 34.30°C | 32.20°C |
[graph 3106]
[graph 3104]
Once we crank up the load on the CPU to its full 95 watts however, the differences start to appear.
The Kabuto and Freezer Xtreme Rev. 2 perform the worst, although the temperatures of both don't create any cause for concern. Of the two, the Freezer Xtreme's advantage, once testing variances are taken into consideration, is negligible.
The H50, MUX-120, and CNPS10X Flex with two fans, all demonstrate a clear advantage over the Kabuto and Freezer Xtreme of at least three degrees. The MUX-120, considering its push-pin retention mechanism, is performing remarkably well and confirms why Intel chose to include it in the LGA1156 press kits. There is a small performance advantage to adding a second fan on the CNPS10X Flex, however at least with this heat-load, certainly not worth the extra expense to do so.
The H50 acquits itself well and clearly demonstrates why water-cooling is a more effective option than even some of the best heatsink-and-fan combos.
| System performance - Idle (Overclocked) | ||||||
|---|---|---|---|---|---|---|
| Cooler | Arctic Cooling Freezer Xtreme Rev. 2 |
Corsair Hydro Series H50 |
Scythe Kabuto | Thermalright MUX-120 (push-pins) | Zalman
CNPS10X Flex (1-Fan) |
Zalman
CNPS10X Flex (2-Fans) |
| Ambient temperature | 23.4°C | 23.6°C | 23.7°C | 23.3°C | 23.6°C | 23.5°C |
| Case temperature | 25.9°C | 26.3°C | 26.7°C | 25.8°C | 26°C | 25.1°C |
| Ambient/Case delta | 2.50°C | 2.70°C | 3.00°C | 2.50°C | 2.40°C | 1.60°C |
| CPU temperature | 39°C | 35°C | 37°C | 38°C | 38°C | 36°C |
| CPU/Case delta | 13.10°C | 8.70°C | 10.30°C | 12.20°C | 12.00°C | 10.90°C |
| CPU/Ambient delta | 15.60°C | 11.40°C | 13.30°C | 14.70°C | 14.40°C | 12.50°C |
[graph 3107]
[graph 3109]Even at idle, the heat load in our overclocked configuration is still substantially higher than at stock speeds. This is because adjusting the CPU multiplier and voltage on the P7P55D Premium seems to disable C1E and EIST, therefore the CPU is always running at maximum clock speed and voltage.
When idle and overclocked, the Freezer Xtreme lags behind marginally from the MUX-120, and CNPS10X Flex with one fan installed. As with the stock-clocked performance when under load, the H50's water cooling is demonstrating a clear lead, and there's still only a small performance delta between the single and dual-fan configurations of the CNPS10X Flex.
| System performance - Load (Overclocked) | ||||||
|---|---|---|---|---|---|---|
| Cooler | Arctic Cooling Freezer Xtreme Rev. 2 | Corsair Hydro Series H50 |
Scythe Kabuto | Thermalright MUX-120 (push-pins) | Zalman
CNPS10X Flex (1-Fan) |
Zalman
CNPS10X Flex (2-Fans) |
| Ambient temperature | 23.6°C | 24.2°C | 24.2°C | 24.1°C | 24.2°C | 23.7°C |
| Case temperature | 26.9°C | 30.6°C | 33.8°C | 28.2°C | 27.6°C | 27.1°C |
| Ambient/Case delta | 3.30°C | 6.40°C | 9.60°C | 4.10°C | 3.40°C | 3.40°C |
| CPU temperature | 78°C | 70°C | 87°C | 78°C | 76°C | 71°C |
| CPU/Case delta | 51.10°C | 39.40°C | 53.20°C | 49.80°C | 48.40°C | 43.90°C |
| CPU/Ambient delta | 54.40°C | 45.80°C | 62.80°C | 53.90°C | 51.80°C | 47.30°C |
[graph 3108]
[graph 3110]Turning up the hurt once more, the Scythe Kabuto performs horribly allowing the CPU to reach a frightening 87 degrees celsius.
Comparatively speaking, the Freezer Xtreme performs better with this overclocked heat-load than at stock, and performance is effectively the same as the MUX-120. It's interesting to note how the MUX-120 at this heat-load loses its edge, the CNPS10X Flex with one fan now beats it to third place. Yet again the H50 takes top spot, but not by much, as the CNPS10X Flex with two fans is finally able to demonstrate a worthwhile performance advantage over a single fan configuration.
In our opinion, with this heat-load and overclock, only the Corsair H50 and Zalman CNPS10X Flex with at least a medium airflow fan (approx. 55 CFM / 94 M3 / hr) are able to safely cool the CPU.