Viewing page 10 of 13 pages.
Previous 1 2 3 4 5 6 7 8 9 10 11 12 13 Next
CASE COMPARISON (continued)
Comparison Two
As configured, the SLK3000B was able to cope reasonably well with the high-end
system used in our testing. Because the power supply was the only thermally
controlled fan in both systems, the excellent behavior of the Seasonic S12-430 kept
the noise level fairly low, even with the system under full load.
For the second comparison, the power supply in each system was swapped for
an older model, a Seasonic Super Silencer 460 with the classic single
80mm fan design. The Super Silencer spent some time as an SPCR favorite, but
it is no longer in production and its noise performance under load is not
on par with the S12-430 used in the first comparison. However,
its acoustics (and fan behavior under load) is still quite good, and representative
of many current quiet power supplies.
Seasonic Super Silencer 460 in the SLK3000B.
|
ROUND 2: Antec P180 vs. SLK3000B, with Seasonic
Super Silencer PSU
|
|
Load
|
System
|
HS Fan
|
CPU
|
GPU
|
AC Power
|
PSU fan RPM
|
SPL
|
|
Idle
|
P180
|
Fanless
|
28°C
|
67°C
|
112W
|
1440
|
26 dBA/1m
|
|
SLK3000B
|
Fanless
|
29°C
|
61°C
|
112W
|
1440
|
27 dBA/1m
|
|
2 x CPUBurn
|
P180
|
Fanless
|
65°C
|
67°C
|
238W
|
1620
|
28 dBA/1m
|
|
SLK3000B
|
Fanless
|
72°C*
|
62°C
|
238W
|
2420
|
38 dBA/1m
|
|
P180
|
Nexus @ 12V
|
54°C
|
67°C
|
239W
|
1620
|
29 dBA/1m
|
|
SLK3000B
|
Nexus @ 12V
|
53°C
|
62°C
|
239W
|
2420
|
39 dBA/1m
|
|
CPUBurn + Prime95 + 3DMark05
|
P180
|
Fanless
|
67°C
|
103°C**
|
294W
(peak)
|
1700
|
29 dBA/1m
|
|
SLK3000B
|
Fanless
|
73°C*
|
92°C**
|
294W
(peak)
|
2900
|
44 dBA/1m
|
|
P180
|
Nexus @ 12V
|
56°C
|
101°C**
|
295W
(peak)
|
1700
|
30 dBA/1m
|
|
SLK3000B
|
Nexus @ 12V
|
58°C
|
92°C**
|
295W
(peak)
|
2900
|
44 dBA/1m
|
|
*CPU Throttling occurred in these configurations.
** VGA temperatures under 3DMark05 correspond to the peak level
seen between separate 3DMark05 benchmarks, and were not easily repeatable.
Temperatures are reported as estimates, and cannot be directly compared.
|
It is interesting to note that the total AC power draw of the system hardly
changed. At full system load, this Super Silencer sample has the about same efficiency
as the S12-430 sample used before. The fan behavior and noise are
a completely different story.
Unlike the S12-430, whose fan hardly ramped up at all during all the testing,
the Super Silencer 400 responded audibly to increased temperature under load.
This was the expected result of swapping the power supply, but it is worth
pointing out that the fan speed increased slightly even in the P180.
CPU Cooling
The CPU and VGA temperatures in the P180 should have been identical to the previous set
to tests. Swapping the power supply in the lower chamber should have no effect on temperatures
in the upper chamber. However, a 1~2°C drop in CPU temperature under load
and a 1~2°C rise in GPU temperature was observed with the different power
supply. The most likely reason for this variance is a slight difference in
the way the cables were routed.
The change in power supply had a greater effect on the CPU temperature in
the SLK3000B. Without the fan on the heatsink, the CPU became hot enough for
protective throttling (of CPU clock & voltage) to occur. This was a 6~7°C
increase, and it would have been greater had CPU throttling not kicked in.
It occurred in spite of the substantial increase in the PSU fan speed, from
1400 RPM at idle to 2400 at CPUBurn load and 2900 RPM at the maximum possible
load. Intuitively, the extra airflow should help CPU cooling, but perhaps
the interaction between the exhaust fan, the side "blowhole" over
the CPU, and the increased fan speed in the PSU reduced the total air speed
over the Ninja heatsink.
With the Nexus fan on the heatsink, CPU temperatures in the SLK3000B were
much better and remained on par with the P180, although they did vary slightly
from the results seen in the first test.
VGA Cooling
No significant changes in the GPU temperature occurred after the power supply
swap. The performance in the SLK3000B continued to be slightly better than
in the P180. The largest difference was observed in the P180 with the GPU
under load, but there was no rational reason for this temperature to change
so much, so the variance is most likely a indication of how unreliable the
GPU load measurements are.
Noise
After the swap to the Seasonic Super Silencer, the overall noise of both
system went up a bit at idle. The change was bigger with the SLK3000B, where
the rougher, sharper quality of the 80mm fan in the Super Silencer could be
distinguished. It was more muted in the P180.
With the CPU under load, however, the difference in noise between the two
systems went from "close" to "no comparison".
While the SLK3000B topped out at 44 dBA/1m under the heaviest load, the P180
stayed at a modest 30 dBA/1m — not silent, but hardly noisy, especially
for a system of this power. In the SLK3000B, the power supply fan could often
be heard ramping up and down in response to changes in heat at its intake,
especially during the dynamic load of the 3DMark05 benchmark. This ramping
was much less marked in the P180, where the increase in noise was gradual.
(Note: There was no point in making recordings of the two cases; one
is LOUD, the other is quiet.)
The lower noise of the P180 system can be attributed to the different position
of the power supply. The separate power supply chamber ensures that the intake
temperature at the PSU remains constant regardless of the heat in the upper
chamber. The PSU only has to deal with its own internal heat, and therefore
its fan doesn't need to spin as quickly or loudly. The choice of power supply
with the P180 is therefore much less crucial than with a more conventional
case.
Comparison Three
The three drives added to the P180 system during
the hard drive testing were installed in the SLK3000B along with the system
drive. The Raptor in the P180 and the Caviar SE in the SLK3000B are close
enough in their level of vibration that they can be counted as identical for
the purposes of this test. (This test was conducted with the S12-430 PSU in
the systems.)
|
ROUND 3: Antec P180 vs. Antec SLK3000B, w/ Four Hard
Drives at Idle
|
|
System
|
System State
|
HDD Temp.
|
SPL
|
|
SLK3000B
|
Idle
|
38°C
|
28-31 dBA/1m
|
|
2 x CPUBurn
|
38°C
|
29-33 dBA/1m
|
|
P180: PSU/HDD Setup 2
(no fan, exhaust vents sealed)
|
Idle
|
43°C
|
26 dBA/1m
|
|
2 x CPUBurn
|
44°C
|
26-27 dBA/1m
|
|
P180: PSU/HDD Setup 4
(fan set to Low)
|
Idle
|
37°C
|
28 dBA/1m
|
|
2 x CPUBurn
|
35°C
|
28 dBA/1m
|
The SPL measurements make clear the difference between the P180 and the SLK3000B:
The three extra hard drives add as much as 6 dBA/1m to the system noise in
the SLK3000B, while adding only 1-2 dBA/1m in the P180. The difference in
even more marked subjectively. The SLK3000B pulses with sound
as the resonances of the different hard drives interact with each other; the
result is a low frequency hum that rises and falls in volume over a period
of ten seconds. The variable nature of the sound makes it hard to ignore.
The P180 didn't exhibit a trace of this sound; even with four drives installed,
no significant resonance could be heard.
MP3:
P180 "Hot Potato", HDD/PSU Setup 2 (four hard drives installed):
26 dBA/1m
MP3:
P180 "Hot Potato", HDD Setup 4 (four hard drives installed, PSU
chamberfan on Low): 28 dBA/1m
MP3:
SLK3000B with four hard drives installed: 28-31 dBA/1m
Comparison Four
As a final proof of the superiority of the drive mounting system
in the P180, the two systems were reverted to their original configurations,
and measurements were made with the WD Raptor hard drive under heavy use. A large
file copy was set up to ensure that the seek noise remained fairly constant
while measurement was taking place. The 10 dBA/1m difference between the two
systems is quite telling, but the most dramatic difference can be heard in
the MP3 recordings of the seek noise.
In the P180, seeks sound much as they do in free air: They are
short and sharp, although the composite panels of the P180 manage to take
some of the bite out of their attack. On the other hand, the SLK3000B does
nothing for the quality of the seek noise: The whole case is shaken with vibration
and resonance when the drive is put to work.
MP3:
P180 "Hot Potato", WD Raptor Drive, idle/seek: 25/29 dBA/1m
MP3:
SLK3000B, WD Raptor Drive, idle/seek: 26/39
dBA/1m
|
ROUND 4: Antec P180 vs. Antec SLK3000B, Hard Drives
Seek Noise
|
|
System
|
System State
|
SPL (Peak)
|
|
SLK3000B
|
Heavy HDD Seek
|
39 dBA/1m
|
|
P180
|
Heavy HDD Seek
|
29 dBA/1m
|
| Help support this site, buy the Antec P180 Black mid-tower case from one of our affiliate retailers! |
|
