Scythe Ninja Cu put to the test: Limited copper edition of the classic
After more than a year of largely unchanged cooler tests, it is time to make certain adjustments to the approach and the tangibility of the tests. The focus is initially on the completely redesigned test system, which will replace the extremely reliable, hot-headed Pentium XE 840 platform. The heating element and heart of the new, current platform is the Intel Core 2 Extreme QX6700, which is also excellently scalable with a specified Thermal Design Power (TDP) of 130 watts, an impressive, powerful quad-core CPU in 65 nm design with four cores each with 2.66 GHz can be operated at a core voltage of 1.35 volts.
As another key component of ours on the quietest possible operation With the power supply, we rely on the Be-Quiet Straight-Power with 600 watts, which was kindly provided to us by Be-Quiet and is available from Caseking , for example is. For reasons of a fair cooler comparison, we exchanged the temperature-controlled, very quiet series fan of the highly efficient power supply for a Scythe S-Flex that rotates at a constant 800 rpm, which replaces the Papst F2GLL as a reference fan in the new system. Another S-Flex also works on the back at the level of the processor socket as the only case fan in the new test case, the Coolermaster Stacker 830, which was made available to us by Caseking for this purpose. The Stacker 830 asThe best-selling high-end housing in Europe offers the convenience of a widespread interior layout, generous space and, thanks to the ample use of perforated sheet metal side walls, good ventilation and air intake options for the various test cooler concepts. All other components can be found in the following list:
- Intel Core 2 Extreme QX6700 'Kentsfield' (4 x 2.66 GHz, 130 watt TDP)
- XFX nForce 680I SLI (ISH9)
- 2x 512 MB Corsair CM2X512A-5400UL CL3
- Asus Radeon X1900XTX [cooled with Arctic Cooling Accelero S1 & Turbo modules @ 5 volts]
- Be-Quiet Straight-Power 600 watts [with Scythe S-Flex SFF21D at 800 rpm]
- Coolermaster Stacker housing [with Scythe S-Flex SFF21D at 800 rpm housing fan]
- Seagate Barracuda V 120 GByte SATA
- Room temperature 23 ° C
- Arctic-Cooling MX-2 thermal compound
There are also some changes in the test procedure with a view to the high-performance quad-core arithmetic. So we give the freshly started system a one-hour warm-up phase, in the Futuremark's 'Ice Storm Fighter' demo (a special focus on the performance ofQuad-core systems coordinated, computationally divided into threads, which heats the entire system accordingly) is used. This is followed by the actual heat test for the processor, whereby we use the small but fine tool Core2MaxPerf 1.0 to keep the CPU load of all four cores constant for another hour. In the test, Core2MaxPerf 1.0 prevailed against some other potential load tools such as Prime95, Intel's Thermal Analysis Tool or S&M due to the particularly intensive thermal stress on all four cores. At the end of the test runs, we give each cooler an idle phase of half an hour. During the entire scenario, the temperature profiles are documented with the Everest Ultimate software and we use the maxima for evaluation. We test each cooler first in 12-volt mode, then in 5-volt mode.
When the CPU is running at full load, our system consumes around 270 watts, while it consumes 155 watts when idling . Attention ! The temperature measurement results are sometimes very much linked to the test platform used and can vary considerably depending on the configuration. They should only be used as a guide and rough tendency, but not as absolute.
The measurement of the volume takes place in individual operation of the respective cooler at a distance of 40 cm with a Voltcraft 320 sound level meter, which measures the noise emissions according to theknown A-weighting curves in a range from 30 to 130 decibels. The basic volume of the measuring room with subjectively perceived silence is 30.4 dB (A). Up to about 35 dB (A), depending on the sensitivity of the user and the frequency character of the sound, it can be assumed that the fan noises are barely audible when the housing is closed. From 40 dB (A) on our scale, the noises can be clearly heard, from 50 dB (A) the very annoying area begins.
As a new service, we are now also offering MP3 audio samples of the cooling combos in the respective operating modes (5 volts/12 volts), since the measured sound pressure values are only really tangible and imaginable for a few readers. In addition, with the help of these recordings, which are made at a distance of 5 cm from the object, the character of the respective fan noise can be better recorded and assessed.
On the next page: Performance comparison