Logic Supply SolidLogic Montivina Fanless Mini-ITX System

Complete|Mobile Systems
Viewing page 4 of 5 pages. Previous 1 2 3 4 5 Next

TEST RESULTS

System Power Draw

Test Results: System Power Consumption (AC)
Test State
GS-L10
Eee Box EB1501
ZBOX HD-ND22
Sleep
2W
2W
1~2W
Idle
19W
20W
23W
x264 Playback
26W
28W
28W
CPU Load
45W
27W
32W
CPU + GPU Load
48W
39W
44W

Though it's running integrated graphics, the GS-L10's mobile Celeron processor lacks SpeedStep capability so the core voltage remains the same regardless of the activity level. As a result, the idle power consumption is only slightly lower than the Atom/ION and CULV/ION combinations found respectively in the Eee Box EB1501 and Zotac ZBOX. On full load it uses just under 50W, high compared to most SFF systems, though it possesses a lot more processing power.

Video Playback

As we've proven before, the GMA 4500MHD is a very capable video decoder — HD video playback on the GS-L10 was smooth. Our video test suite rendered with 20~35% CPU usage, a little high because we were forced to use XP due to drive space limitations; HD decoding is less demanding in Windows 7. A larger SSD with Windows 7 is highly recommended.

Performance Benchmarks

Test Results: Performance Benchmarks
System
Gateway EC1803h
ZBOX HD-ND22
Asus UL30A
GS-L10
CPU
C2S 1.4GHz
Celeron 1.2GHz
C2D 1.33GHz
Celeron 1.9GHz
RAM
3GB
4GB
4GB
2GB
Boot-up
0:53
1:21
0:57
0:37
NOD32
11:14
12:59
11:05
13:00
WinRAR
5:38
6:00
5:27
5:19
iTunes
8:46
9:07
8:02
5:56
TMPGEnc
20:01
12:54
10:22
9:15
PCMark05
2214
3322
3452
4591
3DMark05
979
3686
1396
1206
3DMark06
603
1787
822
728

The combination of XP and a SSD gives the GS-L10 a distinct advantage in boot time over CULV systems, and its Celeron 1.9 GHz processor gives it a significant performance boost, allowing it to win most of our timed benchmarks with ease.

Subjectively, the experience was worse than you would assume giving the benchmark results. The system was at times snappy, but also inexplicably lagged, pausing between commands, particularly during multitasking. With adequate RAM and a decent processor, we believe the problem was with the SSD shipped with our sample. Though it is an SLC model, we believe it uses an older controller prone to stuttering.

Thermal Performance

The system was left at various activity states for 45~60 minutes to allow temperatures to stabilize. The average CPU core temperature was recorded using RealTemp and SpeedFan. Exterior temperatures were measured with an infrared thermometer; the highest measurement recorded on each panel are reported.

Operating Temperatures
State
CPU
Exterior
Top
Left
Right
Idle
27°C
31°C
30°C
30°C
x264 Playback
37°C
37°C
37°C
37°C
CPU Load
57°C
48°C
45°C
45°C
CPU + GPU Load
57°C
50°C
49°C
48°C
Ambient temperature: 21°C.

The case is thermally sound for the system it houses, keeping the CPU and external temperatures well within acceptable levels even under heavy load. The heat was fairly evenly distributed between the 3 panels. The exterior heatsinks did get very warm to the touch, but given the hardware inside and its size, the overall results were fairly impressive, especially since the last passive system we reviewed, the mCubed HFX Micro, had bigger dimensions with only a dual core Atom inside. The extra heatpipes that help distribute the heat more evenly on each panel must help in this regard.



Previous 1 2 3 4 5 Next

Complete|Mobile Systems - Article Index
Help support this site, buy from one of our affiliate retailers!
Search: