Articles: Cooling
 

Bookmark and Share

(13) 
Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 ]

Select Models Tested on CPU Cooler

Now let’s try using the obtained results for building real systems. Of course, it doesn’t make sense to test all fans on a CPU cooler. Moreover, there is simply not enough time for that. That is why I decided to take one fan from each segment of this summary diagram where all testing participants are running with the same level of noise. I picked Noctua NF-S12B ULN from the group of fans with the lowest airflow: at 33 dBA it worked at 720 RPM and created 15.05 CFM airflow. From the middle group I picked Akasa White LED working at 900 RPM with the same acoustics but creating 20.20 CFM airflow. And from the leading group I picked Noiseblocker NB-Multiframe MF12-S3HS at 1080 RPM creating 24 CFM airflow. I would like to remind you once again that the fans worked with the same level of noise maintained at 33 dBA.

Since we didn’t test the fans static pressure in our today’s test session, it wasn’t a good idea to check them out on a cooler with a densely packed heatsink array. That is why I decided to choose a cooler that would be optimized for the airflow and would depend on it the most. It is going to be ThermoLab BARAM heatsink that we have already tested before. The fans were installed onto the heatsink with the airflow directed towards the top of the system case.

We ran all tests inside a system case with the removed side panel. The configuration of our testbed was the following:

  • Mainboard: ASUS P6T Deluxe (Intel X58 Express), LGA 1366, BIOS 1606;
  • Processor: Intel Core i7-920, 2.67 GHz, 1.25 V, 4 x 256 KB L2, 8 MB L3 (Bloomfield, C0);
  • CPU cooler: ThermoLab BARAM;
  • Thermal interface: Arctic Silver 5;
  • Memory: DDR3 PC3-12800 3 x 2 GB OCZ Platinum Low-Voltage Triple Channel (Spec: 1600 MHz / 7-7-7-24 / 1.65 V);
  • Graphics card: ZOTAC GeForce GTX 260 AMP2! Edition 896 MB, 648/1404/2108 MHz (~1030 RPM);
  • Disk subsystem: Western Digital VelociRaptor (SATA-II, 300 GB storage capacity, 10,000 RPM, 16 MB cache, NCQ);
  • HDD silencer and cooler: Scythe Quiet Drive 3.5”;
  • Optical drive: Samsung SH-S183L;
  • System case: Antec Twelve Hundred (default 120 mm fans replaced with four Scythe Slip Stream fans at 840 RPM; 120 mm Scythe Gentle Typhoon at 800 RPM installed on the lower front of the case; default 200 mm fan at 400 RPM at the top of the case);
  • Control and monitoring panel: Zalman ZM-MFC2;
  • Power supply: Thermaltake Toughpower XT 850 W (with a default 140 mm fan).

During this short test session we managed to overclock our 45 nm quad-core processor with the multiplier set at 21 and “Load-Line Calibration” enabled to 3.8 GHz (+42.5%) using the weakest cooling system of the testing participants. The nominal processor Vcore was increased to 1.30625 V in the mainboard BIOS.

 The memory voltage was at 1.62 V and its frequency was around 1448 MHz (7-7-7-14_1T timings). All other parameters available in the mainboard BIOS and connected with CPU or memory overclocking remained unchanged (set to Auto).

You can check any of our latest cooling solutions reviews for the detailed description of the testing methodology (the only difference is that we used latest available software versions). The room temperature at the time of tests was around 26.5-27.0 °C.

Let’s see how greatly the fans with dramatically different airflow readings can affect the CPU cooling. First let’s see the results taken with only one fan attached to the heatsink:

Noctua NF-S12B ULN
(720 RPM)

Akasa White LED
(900 RPM)

Noiseblocker
NB-Multiframe
MF12-S3HS
(1080 RPM)

Next we will check out the results obtained with two identical fans:

2 x Noctua NF-S12B ULN
(720 RPM)

2 x Akasa White LED
(900 RPM)

2 x Noiseblocker
NB-Multiframe
MF12-S3HS
(1080 RPM)

And finally the summary chart for more convenient comparison:

Very impressive and illustrative difference, isn’t it? And note that all fans were working at the same level of noise, in other words, by choosing a proper fan you will be able to significantly lower the CPU temperature without losing any of the acoustic comfort. If we go back to the airflow readings, we will recall that Akasa fan generated about 34.2% higher airflow than Noctua at the same level of noise. And as you can see from the obtained results, it greatly affects the CPU temperature. The airflow from Noiseblocker fan, in its turn, is only 18.8% higher than that from Akasa, which translates into smaller CPU temperature difference. In other words, we see evident practical continuity of the results taken with our thermal anemometer. I would like to add that Nexus Basic fan that is about 4% behind Noiseblocker in airflow taken at the same noise level lost only 1 °C (when tested with two fans on a heatsink):

 
Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 ]

Discussion

Comments currently: 13
Discussion started: 06/23/09 10:35:38 AM
Latest comment: 11/29/10 08:58:10 PM

View comments

Add your Comment