Viewing page 4 of 5 pages.
Previous 1 2 3 4 5 Next
|
Completed system powered up and wirelessly networked.
|
THERMAL/POWER TEST RESULTS
Thermal and power tests were run on the assembled system, mostly to confirm Intel's power specs and establish that the case can adequately cool the board.
|
Test Results: Intel D945GSEJT + Morex T1610
|
|
Test State
|
CPU
|
HDD
|
System
Power (AC)
|
|
Off
|
N/A
|
N/A
|
2.4W
|
|
Idle
|
34°C
|
39°C
|
11-12W
|
|
CPU 100%
|
64°C
|
41°C
|
16W
|
|
CPU+GPU 100%
|
65°C
|
41°C
|
18W
|
|
* 22°C ambient room temperature.
* Turning the WiFi radio reduced power by <1W.
* Plugging the Seasonic adapter into AC w/o attaching the output to the
Intel board/system resulted in 0.9W power consumption.
|
Both CPU and HDD temperature look perfectly safe in our relatively cool room. On the back on a monitor in summer, it may run 10°C hotter, but that's still safe, and few users with normal applications will ever subject this system to the kind of extreme stress applied in our lab tests.
What's interesting is that the power figures of this system are substantially lower than what we recorded for the Asus Eee Box B202, which uses the same CPU and chipset. This suggests higher efficiency in the power circuitry of the board, and perhaps of the AC/DC adapter. They are almost as low as the power figures recorded for the Asus 1005HA netbook. There's almost no point comparing it to the Intel D945GCLF Atom board (with desktop chipset and an 80 Plus ATX12V PSU), which drew at least twice the power in all active operational states..
|
Atom system AC power comparisons
|
|
Test State
|
Intel D945GSEJT + Morex T1610
|
Asus 1005HA netbook*
|
Asus Eee Box B202
|
Intel
D945GCLF + SPI SPI220LE PSU
|
|
Off
|
2.4W
|
1W
|
1W
|
1W
|
|
Idle
|
11W
|
9W
|
16W
|
27W
|
|
CPU 100%
|
16W
|
13W
|
22W
|
31W
|
|
CPU+GPU 100%
|
18W
|
16W
|
25W
|
38W
|
|
All systems featured 1~2 GB RAM and a 2.5" 5400rpm
notebook drive
* The LED-lit LCD screen added 1W to the overall power draw.
** Note that the power of the D945GSEJT was measured with our high accuracy
Extech 380803 Power Analyzer; the others were measured using a Kill-a-Watt
or Power Angel — less accurate, especially below 10W.
|
Aside from the power consumption, the compared systems/boards feature virtually identical performance, as the CPU and GPU are the same, with the amount of RAM being the single largest variant. The Intel GMA 950 isn't good for anything 3D or higher than "easy" 720p video playback, at best, while the Atom doesn't have the muscle to handle Flash video playback or Flash-heavy web sites well.
ACOUSTICS
There's little to discuss with regard to the system's acoustics. Since there are no moving parts on the board or in the case, the only source of noise is the hard drive, so your choice there is obviously critical for noise. The Fujitsu supplied as part of the Intel D945GSEJT kit is a reasonably but not exceptionally quiet modern 2.5" drive. In the damped mounting of the Morex T1610 case, the drive measured about 22 dBA@1m or 20 dBA at the 0.6m distance of the ISO7779 defined seated user position. Seek noise pushed peaks about 2 dBA higher. There was some transmission of vibration into the desktop, but it was modest.
Quieter 2.5" drives are available, for 3-4 dBA lower noise. For a silent ultra small PC these days, a decent sold state drive is the preferred option. In any case, mounting the system on the back of a monitor easily reduces the perceived subjective noise, and drops the measured SPL by 3-4 dBA depending on the environment.
| Help support this site, buy from one of our affiliate retailers! |
|
