Testbed configuration and testing methodology
We have tested both coolers in a closed
system with the following configuration:
·
Mainboard: Intel Siler
DX79SI (Intel X79 Express, LGA 2011, BIOS 0537)
·
CPU: Intel Core
i7-3960X Extreme Edition, 3.3 GHz, 1.2 V, 6 x 256 KB L2, 15 MB L3
(Sandy Bridge-E, C1, 32 nm)
·
Thermal interface: ARCTIC MX-4
·
Graphics card: Gigabyte GeForce GTX 650 Ti 2 GB
(GV-N65TOC-2GI)
·
System memory: DDR3 4 x 4GB Mushkin Redline
(Spec: 2133 MHz / 9-11-10-28 / 1.65 V)
·
System drive: Crucial m4
256 GB SSD
·
Drive for programs and games: Western Digital
VelociRaptor (300GB, SATA-II, 10000 RPM, 16MB cache, NCQ) inside Scythe Quiet
Drive 3.5” HDD silencer and cooler
·
Backup drive: Samsung
Ecogreen F4 HD204UI (SATA-II, 2 TB, 5400 RPM, 32 MB, NCQ)
·
System case: Antec
Twelve Hundred (front panel: three Noiseblocker
NB-Multi-frame S-Series MF12-S2 fans at 1020 RPM; back panel: two Noiseblocker
NB-BlackSilent PRO PL-1 fans at 1020 RPM; top panel: standard 200 mm
fan at 400 RPM)
·
Control and monitoring panel: Zalman ZM-MFC3
·
Power supply: Seasonic SS-1000XP Active PFC F3
1000 W (with a default 120 mm fan)
For our tests, we
overclocked 6 processors with the starting clocking frequency set at 125 MHz,
multiplier of 34x and “Load-line calibration” activated up to 4.25 GHz. The
nominal processor Vcore was increased to 1.35 V in the BIOS mainboard. After
that, we checked the new cooler at even higher the frequency settings and
voltage. The Turbo Boost was disabled during the tests, and the Hyper-Threading
technology was enabled to increase the heat spreading. The memory voltage was
at 1.65 V and its frequency was 2,000 MHz with the time period of 9-11-10-28.
All of other parameters are available in the BIOS mainboard and related to the
CPU or unchanged memory overclocking.
All of the tests were operated on the
Windows 7 Ultimate x64 SP1 OS. We used the following software during our tests.
·
LinX AVX Edition
version 0.6.4 – to load the processor (memory - 4500 MB, Problem Size – 24234,
two 11-minute cycles)
·
Real Temp GT version
3.70 – to monitor the temperatures of the processor core
·
Intel Extreme Tuning Utility
version 3.1.201.5 – to monitor and visually control all system parameters
during the overclocking process
Therefore, given
is the picture describing the whole test session.
The
screenshot
The CPU was loaded by two continuous LinX
AVX testing runs with the setting mentioned above. The stable period for the
CPU’s temperature between two testing cycles is 8 to 10 minutes. We took the
highest temperature recorded to sketch the graph. Moreover, we also provide a
table with the recorded temperature for all cores including their average
values. The surround temperature was checked beside the system case with an
electrical thermometer with the accuracy of 0.1°C, supporting the temperature
changing process every 6 hours. The room temperature during the test fluctuated
from 21.6 to 22 °C.
The noise level of each cooler was measured
between 1:00 to 3:00 pm in a closed room which was 20m2 using an
electronic noise level measurement CENTER-321. As the result, the only noise
sources were the cooler and its fan. Installed on a tripod was a noise meter,
which was kept 150 mm far away from the fan rotor of the cooler. The tested
cooler was located at the edge of the table, on the polyurethane foam. The
lowest noise recorded by our noise meter was 29.8 dBA and the most comfortable
noise level, subjectively, in this testing condition was about 36 dBA (without
mixing with the lower noise level). The rotation speed was adjusted within the
regulated range, using our controller, by changing the voltage with 0.5°C
increment.
We inteded to compared the cooling
performance and noise level of Scythe Katana 4 and Thermalright TRUE Spirit 90
with Thermalright TRUE Spirit 120 which has one default fan:
The
Thermalright TRUE Spirit 120
Performance
Cooling
performance
Following are the results of our tests,
displaying via table and graph. Unlike any of our previous evaluation, they are
classified into three temperature categories for easier analysis: below 80°C,
from 81°C to 90°C and above 90°C.
Testing
results
The
graph
It should be kept in mind that such compact
coolers with an affordable price from Scythe and Thermalright encounter with a
6-core overclocked CPU, are already an achievement. Of course, $25-30 cooler
like Katana or TRUE Spirit 90 is almost not a solution for a $1000 CPU to
overclock it, but our testing results suggest that they can cope with any other
CPU which spreads less temperature like Intel Core i7-3960X Extreme Edition
overclocked up to 4.25 GHz at 1.35 V. That is really an impressive performance.
The differences between Katana 4 and TRUE
Spririt 90 are relatively minor. At the peak speed of the fan, the maximum
temperatures of both coolers are similar to each other (79°C). At 2,000 RPM,
the Katana 4 is 3°C better. The difference is cut down to 2°C at 1600 RPM, but
the TRUE Spirit 90 becomes the leader with 1°C better at 1,400 and 1,200 RPM.
Therefore, two coolers are almost the same in terms of performance. When their
92 mm operate at the speed below 1,200 RP, the coolers cannot manage the
overclocked 6-core CPU.
Compared to the Thermalright TRUE Spirit
120, Katana 4 and TRUE Spirit 90 is $8-10 cheaper, but the maximum CPU
temperature is 7-13°C higher, depends on the fan’s speed. It is your decision
whether the difference in price offsets the difference in the temperature.
Noise
We measured the noise of each cooler
throughout the entire speed range of their fans, and here are the results:
Noise
measurement
Opposite to the previous
testing performance, Scythe Katana 4 and Thermalright TRUE Spirit 90 are
clearly different in terms of noise level. The Katana 4 still keeps its
subjectively comfortable feeling when its 92 mm fan rotates at the speed of
1650 RPM while TRUE Spirit 90 just keeps that comfort at the speed of 1,390
RPM. The Katana 4 becomes inaudible at 1,420 RPM or lower and TRUE Spirit 90 at
1,010 RPM only.
The Scythe seems to be better than
Theralright from the same subjective standpoint. The Thermalright creates some
rattling at the speed of 900-1450 RPM. Thermalright should equip the TRUE
Spirit 90 with a better fan, unless it is the only drawback of our particular
sample.
Conclusion
It is proven that Scythe Katana 4 and
Thermalright TRUE Spirit 90 indeed have high performance, keeping our 6-core
overclocked stable with the clocking rate up to 4250 MHz. Therefore, each of
them is potentially suitable not only for normal PC desktops but also for
overclocking configuration, unless you plan to set some overclocking records. It
is quite attractive when considering their compact dimension, weight and
affordable price. Both products feature the wide range of compatibility and
easy installation. Their availability is also not a big problem. TRUE Spirit 90
is smaller but the Katana 4 is better in noise level. However, any of these two
coolers can possibly be an excellent solution for any moderately overclocking
PC. Low price but high quality, Scythe Katana 4 and Thermalright TRUE Spirit 90
is really a bargain.
